Recent Marine Geological Research in the Mariana and Izu-Bonin Island Arcs!

REVIEW ARTICLE

PATRICIA FRYER 2

ABSTRACT: Over the last decade models of the structure, composition, and evolutionary processes of the Mariana and lzu-Bonin island arc systems of the western Pacific have been changed profoundly by an interdisciplinary and multi national approach to studying the details ofthese regions. The standard marine geological studies have been superseded by detailed sea-floor mapping, sea-floor observations using submersibles, and deep-ocean drilling efforts. The increased level of effort is the result of discoveries of new geological phenomena in these regions and the desire to approach the study of ancient convergent margin terrains exposed on land in the light ofthe resultant insights. The new discoveries include active eruption ofserpentine muds forming large volcano-like seamounts in the Mariana forearc, recent protrusion ofserpentinite debris flows from fault dissected horsts of metamorphosed supra-subduction zone mantle also in the Mariana forearc, similar ancient seamounts in the Izu-Bonin forearc region , evidence of recent forearc rifting, petrogenetically complex arc volcanoes situated within the backarc basin setting in both the Mariana and lzu-Bonin systems, the occurrence in close proximity to one another ofmagmas generated from different sources (the deep arc source and the shallower backarc basin basalt source) in neovolcanic zones ofthe incipient rifting portions ofthe Mariana and Izu-Bonin backarc rifts, and a variety ofunique hydrothermal systems and fauna associated both with the arc volcanoes and with the active backarc spreading centers.

THEACTIVE DEEP OCEAN trenches of the west cesses of subduction in an environment unin ern Pacific mark the zones of collision and fluenced by preexisting continental crust. In subduction between the Pacific lithospheric these relatively uncomplicated island arc sys plate and the Asian continent. South ofJapan tems it is simpler to decipher the components this zone of collision takes place between the ofthe process ofplate subduction and to study Pacific plate and the Philippine Sea plate, mak them . Only in this way can we begin to deter ing interoceanic collision zones. The island arc mine how these components are related to one systems generated in this region ofthe Pacific, another and how subduction relates to the the Mariana and Izu-Bonin arc systems , are of evolution of the earth's lithosphere. particular interest to marine geologists because The physical effects of subduction on the such convergent plate boundaries provide an wedge of cru st and mantle overlying the de opportunity to study several fundamental pro- scending lithospheric plate, in what is now termed the supra-subduction zone, are ofpar ticular concern to those interested in producing models of the nature of interactions between 1 Hawaii Institut e of Geophysics Contribution 2221. Manuscript accepted I July 1989. lithospheric plates. The type ofmodel consid 2 Hawaii Institute of Geophysics, 2525 Correa Rd., ered for each con vergent margin depends on Honolulu, Hawaii 96822. the subduction system. There are two end

95 96 PACIFIC SCIENCE,Volume 44, April 1990

member types ofconvergent margins in the in on a small scale than is that of the seamount teroceanic settings, accretionary and erosional. poor plate being subducted beneath the Izu Accretionary systems are those in which Bonin system. Input to the Mariana arc system portions of the down-going plate are detached, from constituents driven off the down-going usually including part ofthe sedimentary col plate may therefore be more heterogeneous. umn and sometimes part of the underlying The line of active volcanoes in the arc is igneous basement. These pieces of the sub known as the volcanic front. Behind the vol ducting plate are then accreted as slivers, canic front, in many convergent margins, such wedges, or en masse as exotic terranes onto the as the Mariana and Izu-Bonin systems, rifting toe of the overriding plate . Erosional systems, of the arc produces young ocean basins . As such as the Izu-Bonin system, are those in the basins evolve they progress from a series of which the toe of the overriding plate is me rift graben to a system of volcanic rift seg chanically eroded by normal faulting as the ir ments offset by transform faults. The backarc regular surface of the down-going plate passes basin regions of the Mariana and lzu-Bonin beneath it. Although it is generally accepted systems are particularly useful because they that both types of convergent plate margins can provide information regarding the tec currently exist within the ocean basins, there is tonic evolution of the backarc basins from the growing acceptance of the concept that a given earliest stages of rifting (both Mariana and convergent margin may change from one mode Izu-Bonin) to the more fully evolved (Mari to the other with time. A convergent margin ana). The results of the tectonic studies of may even exhibit both modes , varying with these two backarc environments are being local environments along strike. Such a situa compared with studies of continental rifting tion has been suggested for the Mariana sys as well as with studies of midoceanic ridge tem, which is erosional over most ofits length, rifting in an effort to contrast the rifting styles but is possibly accreting slivers of the Pacific of convergent margin settings with those of plate in the southernmost latitudes. To under purely divergent margin settings (the conti stand the various factors that control the na nental rifts and midocean spreading centers) . ture of subduction in any given convergence The backarc basins produce magmas that situation, a detailed model based on field evi are similar to midocean ridge basalts in some dence as well as theoretical considerations regards, but have a unique compositional must be produced and tested. overprint from the arc that distinguishes them Understanding the nature ofthe interactions as a new class ofbasalt. Clearly, as the backarc between the subducted slab and the mantle basins evolve, the structural similarities to through which it descends is critical for pro midocean ridges increase. If the magmas ducing working models of subduction systems. maintain a distinct composition, then even if Of obvious interest among these interactions the structural components of an exposed sec is the generation of magmas in the supra tion through a backarc basin terrane appear to subduction zone environment and the forma represent a midocean ridge setting, it would be tion ofisland arc volcanoes. Study of the evo possible to trace the origin ofsuch an exposure lution of the island arc volcanoes can docu to an arc environment. Geochemical studies ment not only the timing of subduction and of some subaerial suites of rocks that include the structural evolution ofthe arc system, but oceanic sediments, basalts, gabbros, and as also provide a means by which to trace aspects sociated mantle rocks (ophiolite sequences) of the chemical interaction between the down show that these rocks formed in the supra going plate and the supra-subduction zone subduction zone environment, either in a mantle. The Mariana and Izu-Bonin systems volcanic front setting or in a backarc setting. are very different in the types of oceanic plates Volcanological studies of the Mariana and being subducted. The plate being subducted Tzu-Bonin arc volcanic fronts and backarc beneath the Mariana system contains numer basin spreading centers provide the oppor ous large seamounts. Thus, the composition tunity to investigate eruptive processes, both of the plate is likely to be more heterogeneous subaerial and submarine, and the generation Review Article: Marine Geological Research-FRYER 97 of several types of economical1y valuable de Mariana and Tzu-Bonin systems; oflarge, ac posits, such as porphyry ore deposits, Kuroko tive backarc basin volcanoes erupting arc type deposits, and hydrothermal vent deposits. lavas; of the occurrence in close proximity to Recently, attention has been drawn to a far one another ofmagmas generated from differ less obvious locus for generation ofmagmas, ent sources in neovolcanic zones of the inci the region trenchward of the volcanic front, pient backarc rifting portions of both systems; termed the forearc. Located between the trench and of active hydrothermal processes in the axis and the line of volcanic islands, the fore arc and backarc regions of both systems pro arc region had long been considered too cold to vided the impetus to perform in situ studies of permit the generation of magmas. Petrologists these regions. A series of deep submersible now recognize the importance ofthe evolution (Alvin) dive cruises and two cruises of the of magma types specific to the forearc envi Ocean Dril1ingProject were completed within ronment (boninites, named for the Bonin the last two years in the Mariana and Izu Islands where they were first described) and Bonin systems to study these new phenomena. have begun to provide detailed studies oftheir This paper documents these latest advances petrogenesis and distribution. To understand in marine geology of the Mariana and Tzu the process of magma production in the supra Bonin arc systems. Itprovides a technical sum subduction zone region , particularly in the mary of the development of ideas concerning forearc , we must investigate the distillation of the tectonic evolution of the arc systems and the descending slab, the flux within the supra summarizes the major findings of the Alvin subduction zone environment ofthe constitu dive cruises and the dril1ing results. ents driven off the slab, the associated meta morphism ofthe supra-subduction zone man Tectonic Setting tle and crust, and the secondary convection patterns set up by subduction that may serve Early investigators ofthe Mariana and Izu to redistribute the distillates. Bonin systems of volcanic and remnant arcs The dehydration of the down-going slab and backarc basins (Figure 1)provided a vari produces profound effects on the supra-sub ety of evolutionary schemes to account for duction zone mantle. In the earliest stages the current configuration of the major tec of subduction the cooling effect of the down tonic elements of the region. Uyeda and Ben going plate permits chemical changes in min Avraham (1972) suggested that the Philippine eralogy of the supra-subduction zone mantle Sea plate formed as a trapped piece of Kula/ to proceed, producing low-temperature/high Pacific plate crust when reorientation of pressure metamorphism. Once the subduction spreading occurred in the Pacific ca. 42 mil process has evolved to the point at which the lion years ago (mya). They further suggested Mariana and Izu-Bonin systems currently op that the subduction was initiated along a for erate, the potential for metamorphism within mer transform fault offset some 1500 km and the forearc is profound. Virtual1y the entire oriented roughly north-south between two forearc region can be metamorphosed in the east-west spreading center segments of the low-temperature/low-pressure to low-temper Pacific/Kula midocean ridge. Shih (1980) pro ature/high-pressure ranges . This means that vided magnetics evidence that confirmed the vast amounts of the constituents driven off entrapment hypothesis, but noted that the the down-going plate can be accommodated western Philippine basin drifted north ca. 15 within the supra-subduction zone mantle. to 20° and underwent clockwise rotation of After two decades ofincreasingly more de about 50 to 70° since initial formation at about tailed investigations, the study ofthe Mariana 40 mya. Klein and Kobayashi (1980)and Lewis and Izu-Bonin systems has burgeoned; The et al; (1982) later suggested that the northern discoveries of actively erupting serpentine portion of the Philippine Sea plate, from the mud volcanoes in the Mariana forearc region; Oki-Daito ridge to the Palau-Kyushu ridge of uplifted blocks or protrusive diapirs of and the intervening basins, formed as a con metamorphosed forearc mantle in both the sequence of southward convergence, arc for- 98 PACIFIC SCIENCE, Volume 44, April 1990

CHINA

4 0·

PACIFIC 35· PLATE

30·

.: 20· WEST

PHILIPPINE

15·

FIGUR E I. Location map showing major tectonic elements of the western Pacific in the vicinity of the Philippine Sea. Solid lines with solid triangles depict the axes of the trenches bounding the volcanic arcs and backarc basins of the region. Dashed doub le lines represent extinct spreading centers in the backarc basins. Narrow, solid double lines represent the axis of spreading in the active backarc basin-the Mariana Trough and the Bonin Rifts. Review Article : Marine Geological Research- FRYER 99 mation, and episodes of backarc rifting. They were collected from a deep fault scarp nearly suggested that this region was moved north 70 km from the trench axis (Johnson and ward as an integral part of the Philippine Sea Fryer 1988). The occurrence ofMORB in these plate during the translation/rotation episode. dredges requires a reevaluation ofthe existing From ca. 21 to 15 mya the Parece Vela and models of evolution of the forearc regions. Shikoku basins were opening, and the separa Further work will be required to determine tion of the Mariana/Izu-Bonin arcs from the whether these MORB lavas represent a frag Palau-Kyushu remnant arc occurred (Scott et ment ofold oceanic plate or indicate episodes al. 1980). The opening ofthe Mariana backarc of forearc rifting and generation of supra basin began at ca. 6.5 mya (Hussong and subduction zone ophiolites. In addition to the Uyeda 1981), and the initiation of the Izu arc volcanics, there are numerous exposures Bonin rifts probably began as recently as I of ultramafic rocks in the outer portions of mya (Brown and Taylor 1988). both the Mariana and the Izu-Bonin forearcs The nature of the forearc regions, those (Bloomer 1982, Fryer et al. 1985a, Ishii 1985, areas between the volcanic front and the axis Bloomer and Hawkins 1987, Fryer and Fryer of the trench (Figure 2), has provoked consid 1987). Large seamounts of serpentinite are erable debate over the last two decades . The forming over a zone ca. 100 km wide on the recognition that accretionary and erosional Mariana forearc (Fryer et al. 1985a, Hussong processes both can be active in supra-subduc and Fryer 1985, Fryer and Fryer 1987), and tion zones and can produce vastly different older, sediment-draped seamounts are exposed types of forearc regions is a major advance in on a narrow ridge on the inner trench wall of our understanding ofthe evolution of forearc the Izu-Bonin system within 50 km of the terrains (Karig 1971, Hussong and Uyeda trench axis (Honza and Tamaki 1985, Ishii 1981,Karigand Ranken 1983,Fryer and Fryer 1985, Fryer and Fryer 1987). The origin of 1987). Both the Mariana and Izu-Bonin fore these seamounts, first described as serpentinite arcs seem to be primarily erosional. No large diapirs by Bloomer (1982), is currently under accretionary wedge has been noted in either of study. The seamounts vary from horst blocks these forearc regions, and the results ofstudies of uplifted metamorphosed ultramafics to of rocks collected from these areas indicate structures similar to mud volcanoes , but with that they formed primarily of arc magmas far larger edifices, 20 km in diameter and 1500 (Skornyakova and Lipkina 1976, Detrich et m high. The nature of the active eruption of al. 1978, Meijer 1980,Meijer et al. 1981 , Wood serpentine muds, the composition of associated et al. 1981, Bloomer 1982, Beccaluva et al. fluids seeping through the vent systems, and 1980, Bloomer and Hawkins 1983, Fryer et al. the composition of the chimmey structures 1985a, Honza and Tamaki 1985, Ishii 1985, associated with the seeps are unique in the Fryer and Fryer 1987, Johnson et al. 1987, ocean basins (Fryer and Fryer 1987, Haggerty Saboda et al. 1987). Even in the deepest expo 1987b). The serpentinite seamounts provide sures of the inner trench walls the predominant an opportunity to study deep-seated processes rock type is an island arc basalt. Only very ofsupra-subduction zone metamorphism and small volumes of alkalic basalts, presumed to fluid flux. The mud volcano -like seamounts be off-scraped fragments of oceanic islands, apparently entrain country rock from the walls were noted in the samples recovered close to of the conduit and thus sample deeper-level the inner wall ofthe Mariana trench (Bloomer rock types than can be reached even by deep andHawkins 1983). Thus, it was suggested sea drilling techniques. that the forearcs had been formed by arc vol In an effort to integrate existing geophysi canism and that the arc volcanics were exposed cal and petrochemical data and to provide a in the inner trench wall oy fe'Cto iiic erosion regional geologic framewoiYfor sliiClies ofthe (Bloomer 1982). Recent dredge samples from detailed geology ofthe Mariana and Izu-Bonin the midregion ofthe Mariana forearc provide arc systems, a series of 12 SeaMARC II side a different scenario. Basalts with trace element scan sonar and bathymetry surveys was con signatures of midocean ridge basalt (MORB) ducted in the Mariana and Izu-Bonin regions 100 PACIFIC SCIENCE, Volume 44, April 1990

A B 400N ,....---~-r---r---_33°N .....,~~....---.....,.-....,.--.--,...~~~...,.,.,.r-,.,.,

10° ...._-""_...w.;o__....._ ... 130 0E C 21°N ,....--~-..,...... ,.-'""\"""'I""I"r....,-,...... -..,...-~.....~---.,...----__.

19° V 6 ~

18° 145° E 146° 147° 148° 149° 150°

FIGURE 2. Maps of the major tectonic elements of the Philippine Sea region, the Mariana and Izu-Bonin arcs: A, a general locator map, showing locations of the detailed bathymetry maps (shaded areas) that are parts Band C of this figure (after Fryer and Pearce et al. 1989b); B, detailed bathymetry map ofthe Izu-Bonin forearc region contoured in IOOO-m intervals, showing locations of Ocean Drilling Program (ODP) leg 125 drill sites; C, detailed bathymetry map of the Mar iana forearc region contoured in 1000-m interval s, showing locations of ODP leg 125 drill sites. Review Article: Marine Geological Research-FRYER 101 over the past 8 yrs. The results ofthese surveys volcanism and hydrothermal activity in the define neovolcanic zones in the northernmost arcs and in the backarc rift basins of these Mariana backarc basin (Beal 1987) and in the systems. The Mariana arc system was the pri Sumisu rift (Brown et aI. 1988, Taylor et aI. mary focus of the dive cruises: five of them 1988). The surveys show the diversity of lava concentrated on the Mariana system. Three of flow fields and structures associated with the cruises dealt with petrology, tectonics, and backarc spreading centers and arc volcanoes hydrothermal systems of the Mariana backarc of the northern Mariana arc (Hussong and basin, the actively spreading, young ocean ba Fryer 1983, BeaI1987). In addition, the struc sin immediately west of the active islands of tural evolution of the innermost part of the the Mariana arc . The first dives were on the northern Mariana forearc was investigated active spreading center in the basin at about using side-scan and bathymetry data together 18°13' N (Figure 3). One cruise studied the with seismic reflection data (Mahoney and serpentinite seamounts of the Mariana forearc. Fryer 1988) and indicated a recent change in One focused on the volcanoes ofthe northern stress regime in the northern part of the Mari Mariana arc . One dealt with the arc and back ana forearc. Structures there suggest that fore arc volcanic centers of the central Izu-Bonin arc rifting may playa significant role in the arc system, the Sumisu rift. development of an arc system at any stage in its evolution. Investigations of the outer part MARIANA BACKARC BASIN STUDIES AT 18°13'N. of the Mariana forearc show the diversity of The first of the dive cruises to investigate the types of serpentinite seamounts (Hussong and rift had been prompted by the discovery of 3He Fryer 1985) and indicate the nature of the CH 4 / plumes in the water column 800 m processes that favor their formation (Fryer above the sea floor (Horibe et aI. 1986). The and Hussong 1985, Fryer and Smoot 1985, dives on the rift axis revealed a series of axial Fryer et aI. 1985a, Fryer and Fryer 1987). volcanoes and two large hydrothermal fields With adequate background information and on the flanks of two of these volcanoes at detailed field mapping accomplished, it was depths of 3600-3700 m (Craig et aI. 1987, possible to design studies using the Alvin sub Kastner et aI. 1987). However, the vent waters mersible and deep-sea drilling cruises to inves are similar to those of Loihi Seamount (the . tigate specific aspects of the geological devel active volcano immediately south ofthe island opment of the Mariana and Izu-Bonin inter of Hawaii) vents in 3Hej4He ratio (8.6 x at 3He oceanic arc systems. Specific details of the in mospheric [R/RA = 8.6]), with CH 4 / = 6 dividual Alvin cruises may be found in Fryer et (0.5 ± 0.2) x 10 , and thus are very different 3He aI. (in press a,b), Hochstaedter et aI. (in press), from the plumes (R/R A = 3.2, and CH 4 / 6 Johnson and Fryer (in press), and Urabe and > 100 x 10 ) that had led to their investiga Kusakabe (in press). More comprehensive tion (Craig et aI. 1987). Several smaller fields summaries of the drilling cruises may be found were discovered on the second dive cruise in Fryer and Pearce et aI. (1989a,b) and (Hessler et aI. 1987). The hydrothermal vents Fujioka and Taylor et aI. (in press). were variable in terms of temperature, ranging from '" 6°C to a maximum of 287°C. The high-temperature chimney structures produce Recent Alvin Submersible Studies "clear smokers" with vent water temperatures A series of six Alvin submersible cruises of up to 287°C in contrast to the " black took place from April through August of 1987 smokers" ('"350°C) that have been discovered in the Mariana and Izu-Bonin interoceanic on midocean ridge hydrothermal systems arcs. The main objective of these cruises was (Campbell et aI. 1987). The structures are to study in detail the geologic evolution of dominantly composed ofbarite and silica, but these arc systems of the western Pacific. The also contain spalerite, galena, and minor chal specific proce sses outlined for study ranged copyrite (Craig et aI. 1987, Kastner et aI. 1987). from active serpentinite diapirism in the As compared with midocean ridge vents of forearc environment to active and incipient similar temperatures, the composition of the 100 K I LOMETERS

' l.A XIA L RIFT VALLEY AND SPREADING AXIS \ "'\\ [JDETA ILED SURVEY AREA (FIG.2) DEEP SEA DRILLING rTl.':,. >4 km * CORE SITE L.lJ

14~OE 1440 E 145°E 146°E I I I

F IGURE 3. Bath ymetry map of the central part of the Mariana backarc basin contoured in 500-m intervals (after Lonsdale and Hawkins 1985), showing location of the spreading ridge (dark hatchured lines), location s of Deep Sea Drilling Project (DSDP) leg 60 drill sites, and location s (boxes) of the Alvin dive regions on both the Mariana backarc basin spreading center (right) and on the off-axis, hydrothermal mounds area (left). Depths greater than 4000 m are shaded. Review Article : Marine Geological Research-FRYER 103

Mariana vent fluids and chimney structures is tural variation in the volcanic centers are an different from those of hydrothermal regions important component in the interpretations on midocean ridge spreading centers (Camp of both the distribution of the hydrothermal bell et al. 1987, Kastneret al. 1987).The pH of fields and the petrogenesis of the lavas recov the water samples from the Mariana high ered during the dives. Thus, as a consequence temperature vents is 4.39 compared with 3.8 of the second dive cruise the structure of the for midocean ridge vent fluids, and the alka rift volcanoes and relationships of the struc linity of the Mariana vent waters is also higher, ture to the petrology of the lavas distributed 0.43 meq /I compared with - 0.19 meq /l for along the rift were integrated with more de midocean ridge vent waters (Campbell et al. tailed investigations ofthe hydrothermal fields. 1987). Boron in the vent waters oftheMariana The dives confirmed the remarkable variation chimneys is 200% that of ambient seawater in structure along the rift that had been dis and has a b 11 B of +20/mil, whereas midocean covered by multibeam surveys (Hawkins et al. ridge vents are usually only 10to 20% enriched 1987). A total of 18 Alvin dives augmented by relative to seawater, with a b 11 B of +26 to 26 dredges detail the composition of the basalts +32/mil (Campbell et al. 1987). It appears and more evolved lavas that were erupted from from both composition of the chimney struc the largest edifices (Hawkins et al. 1987). The tures and from the composition ofthe Mariana lavas have a source similar to that of MORB vent fluids that precipitation of sulfides must based on the major element compositions be occurring at depth within the ridge crest (Hawkins et al. 1987). This observation con hydrothermal system, although no samples of firms earlier work based on dredge and deep such precipitates were recovered (Campbell et sea drilling samples (Fryer et al. 1981, Wood al. 1987, Kastner et al. 1987). et al. 1981 , Sinton and Fryer 1987), but pro Biological studies of these vents were plan vides far greater detail in terms ofdistribution ned in an attempt to characterize for the first ofvarious rock types along axis. Isotopic anal time the detailed nature of vent communities yses and the Rb and .Sr concentrations of the in the western Pacific. The vent fauna associ basaltic glasses from some ofthe samples con ated with these fields is described by Hessler et firm an intimate mixing of the MORB source al. (1987) as rich and unusual. Depending on with an arclike source component (Macdougal the temperature of the fields, the fauna of the et al. 1987) on a scale not discernable with communities varies. The low-temperature analytical results from dredge samples (Sinton ( ~ 6°C) regions are characterized by abun and Fryer 1987). Thus, the geochemical results dant widespread anemones and clusters of from the Alvin studies provide a far more de barnacles around discrete vent openings. As tailed examination of the relationships of the sociated with the barnacles and anemones are petrology to the structure of the spreading galatheid crabs and polynoid polychaetes. The center. fields with slightly higher temperatures (from The last of the Marianabackarc basin dive ~ 15°C to 250°C) contain clusters ofa unique cruises (the fourth cruise to the Mariana re " hairy" gastropod around the vents, grazing gion) concentrated on the hydrothermal sys bresilid shrimp and brachyuran crabs, encrus tems occurring off-axis in the central latitudes tations ofbarnacles, paralvinellid polychaetes of the basin . Detailed studies of fields of off and associated harpactacoid copepods, many axis hydrothermal mounds, discovered by limpets, occasional mussels, and dispersed heat-flow and deep-tow studies (Leinen and anemones. Anderson 1981,Hobart et al. 1983, Lonsdale The second dive cruise on the spreading axis and Hawkins 1985), their structure, sediment provided structural and petrologic studies of composition, associated heat flow, pore water the rift center volcanoes. These studies are composition, nutrient profiles, and pore pres extensions of deep-tow surveys, multibeam sures were the focus of the Alvin dives in the studies, and dredging work that had been per mounds area (Leinen et al. 1987, Wheat and formed before the diving cruise (Lonsdale and McDuff 1987). The mounds, located ca. 50 Hawkins 1985, 1987). The details ofthe struc- km west of the Mariana backarc rift center, 104 PACIFIC SCIENCE, Volume 44, April 1990

show high heat-flow values exceeding 10 which they suggested indicated incorporation watts /m, and nonlinear concave thermal of sediments into igneous melts, subsequently gradients indicate an active hydrothermal sys superimposed on a backarc basin basalt tem, recharge of the system, and potential substrate. The distribution of the areas of escape of fluids along fault scarps within the high heat flow and the associated mounds region (Leinen and Anderson 1981, Hobart et appeared to be controlled by faulting, not by al. 1983, Leinen et al. 1987). The mounds evenly disseminated hydrothermal activity occur in regions of high heat flow that are (Leinen et al. 1987). The mounds are struc located in several semilinear depressions turally complex and associated heat-flow oriented roughly north-south along a north values indicate variability even within a given south trending ridge. The mounds vary in mound. The vertical pore pressures measured morphology from smooth to pockmarked to in in situ measurements were the highest ever hummocky features 1 to 2 m high. The heat measured in hydrothermal mounds in the flow within the mounds area was monitored ocean basins. The distribution of the mounds by the first in situ long-term heat-flow moni and the local topography confirm the struc tor and was found to fluctuate over a period of tural control over the formation of the mounds 10 days at the vent of the mound studied, but (Leinen et al. 1987). smoothly decreases exponentially away from the mounds (Leinen et al. 1987,Yamano et al. MARIANA FOREARC STUDIES. Serpentinite 1987). Investigations of the nutrient profiles seamounts in the Mariana forearc region have within the areas of high heat flow and of the been particularly well studied at 19°30' N, physical properties of the sediments collected 147° E (see Figure 2). There, side-scan and in push cores from the mounds add informa bathymetry surveys and the results of petro tion critical to the understanding ofthe hydro logic and geochemical studies reveal the pres thermal system active in the mounds area. ence of two distinct forms of serpentine sea Nutrient profiles from piston core data corre mounts (Hussong and Fryer 1985, Fryer and late directly with the pattern of heat flow and Fryer 1987). These two edifices were the sites show generally higher nitrate (constant with ofdives on the third Alvin cruise in the western depth) and silicate (constant or decreasing with Pacific. A young , actively erupting serpen depth) concentrations than bottom water and tinite seamount, informally named Conical that phosphate concentrations decrease or re Seamount, lies at the intersection of several main consistently lower than 1 microMolar, fault lineaments at the edge of a large forearc whereas in areas of low heat flow silica and graben (Newsom and Fryer 1987). It is similar phosphate concentrations increase and nitrate in appearance on side-scan imagery to a mud concentrations decrease with depth (Wheat volcano, the flanks of which are covered with and McDuff 1987). Increasing porosity with long sinuous flow forms. An apparently older, depth in the piston cores suggests that the pore possibly diapiric seamount or uplifted horst water advection driven by the hydrothermal block of serpentine, informally named Pac system disrupts processes of normal sediment man Seamount, is situated to the southeast of consolidation (Dadey and Leinen 1987). Pre Conical Seamount. Pacman Seamount is cut vious studies of sediments from the area by east-west trending faults and is breached revealed hydrothermal mineralization similar on the east flank by a fault-bounded collapse to that encountered in off-axis mounds in the graben. This eastern-flank graben is partially Galapagos region (Leinen and Anderson filled by a large oval-shaped flow of serpen 1981). However, none of the sediments tinite debris. The objectives of the dives on reported from the Alvin dives show chemical these seamounts were to investigate the upper alteration similar to that encountered in the flanks-of the seamounts in the vicinity offlow Galapagos mounds (Leinen et al. 1987). The features observed on side-scan images and to earlier studies of Lonsdale and Hawkins study what active processes occur at the sum (1985) indicated that the mounds area is com mit of the seamounts, to sample the flow ma plicated by an episode of silicic volcanism, terial and any chimney material that might be Review Article: Marine Geological Research-FRYER 105

present on the seamounts, and to determine tropods, and bacterial mats were collected the composition of vent fluids associated with from the chimneys. Twenty-two bacterial iso the seamounts. lates were produced from the samples of the Eleven dives were completed on these two chimneys , and studies oftheir possible genetic seamounts. The dives explored the serpentine interrelationships to the chimneys are in pro flows on the flanks of Conical Seamount and gress (Cloutier and Haggerty 1988, Haggerty the faulted southern flank. The rock types and Cloutier 1988). Fluids actively seeping recovered in dives and dredges from the flanks from one ofthe chimneys were sampled at the of the seamounts reveal a variety of meta active silicate chimney and revealed an unusual morphosed crustal and upper mantle rocks chemistry. The pH of the fluids is high (9.28 (Saboda et aI. 1987). The dredge hauls on the versus 7.72 for ambient seawater), alkalinity is fault scarps surrounding the large forearc also high (5.53 versus 2.41 meq/l), and the graben east of the seamounts have revealed vent waters are enriched in CH 4 (l000 versus the first evidence for MORB composition 2.1 nM) , Si02 (0.75 versus 0.12 mM) , S04 lavas in a forearc setting (Johnson and Fry (30.4 versus 28.6 mM), and H 2S (2.1 mM er 1988). The fault scarp from which these versus not detected in ambient). Note that MORB-like basalts were retrieved is ca. 2000 seawater probably mixed with the fluids dur m high and exposes the MORB-type basalt at ing sampling, because the rate ofseepage from high levels in the outcrop. Thus, the MORB the chimney is very slow. Thus, these enrich lavas are not deeply buried under an arc ments must be considered minima. The en generated carapace, as would be expected if richments noted imply that deep-seated ser the models offormation of the supra-subduc pentinization processes as well as interaction tion zone lithosphere in the Mariana forearc of seawater with crustal rocks and interac as previously espoused are correct. Further tions of seawater with the surficial serpen more , although most of the basalts from tinite may all contribute to the composition of Conical Seamount have island arc trace ele the fluids. The deep-seated origins ofthe fluids ment characteristics, some also have trace ele are substantiated by work on similar chimney ment signatures ofMORB. This indicates that samples collected from a dredge of a serpen the presence of MORB-type lavas can be tinite seamount in the southern portion of the traced over a region at least 40 km wide in the Mariana forearc. The isotopic C and 0 com vicinity of the area studied. The variety of position of the southern seamount chimney rock samples , ultramafic to mafic, collected samples indicates that a source other than from the dive and dredge sites is further con seawater is involved in the generation of the firmation of the high degree oftectonic defor fluids (Haggerty I 987a). Fluid inclusion work mation in the Mariana forearc and of the de performed on aragonite needles from the Alvin gree of faulting associated with formation of samples from Conical Seamount and from the serpentinite seamounts (Fryer and Fryer 1987). southern seamount dredge samples indicates At the summit ofConical Seamount the dives the presence of methane and higher hydro revealed recent flows of unconsolidated ser carbons, the sources for which are probably pentine devoid of sediment cover. In a small deep-seated within the supra-subduction zone region (ca. 100m in diameter) on the west side (Haggerty 1989). of the two summit knolls, chimney structures Dives on Pacman Seamount explored the composed of both carbonate and silicate large oval-shaped flow ofserpentine that par (Fryer et aI. 1987b, Haggerty 1987b) were dis tially fills the graben on the seamount's east covered. The chimneys varied in morphology ern flank, the summit region ofthe seamount, from a corroded appearance in the carbonate and the faulted portion ofthe southern upper (aragonite and calcite) chimneys to a smooth flank. Pacman Seamount showed no signs of appearance in the silicate chimneys (com recent activity, and no chimney structures posed of a new mineral, a Mg-silicate analo were observed on the summit dive. The upper gous to allophane [Haggerty 1987b , Haggerty half of the fault scarp on the southern margin and Cloutier 1988]). Small limpets , gas- of the flank graben on the eastern side of the 106 PACIFIC SCIENCE, Volume 44, April 1990

seamount exposes massive serpentinite with blocky morphology and no flow matrix ma terial. Thus, the interior ofthis portion of the seamount is probably massive serpentinized ultramafic rock exposed by normal faulting within the forearc . Pacman Seamount shows no evidence of being composed of layered flows of the kind observed on Conical Sea mount. The results of the Mariana forearc dives pro vide clear evidence that the serpentinite seamounts can be formed by a variety ofpro cesses, that they occur in association with large-scale faulting on the forearc, and that the serpentinite probably originated from the supra-subduction zone man tIe. Also it is clear that fluids emanating from the active Conical Seamount are at least in part of deep-seated origin and that the serpentinite flows have entrained crustal-level basalts during em placement. These seamounts differ from ser pentinites of midoceanic fracture zones in the chemistry of authigenic minerals with which they are associated (Haggerty 1987a) and in their manner of emplacement (Fryer and Fryer 1987). They likely pro vide in situ examples of processes that are analogous to those that have formed serpentinite deposits and struc tures noted in subaerial exposures of conver gent margin terranes in numerous locations worldwide (Lockwood 1971 , Carlson 1984).

MARIA NA ARC VOLCANISM AT 22° N. The vol canoes of the northern part ofthe Mariana arc (Figure 4) are ofinterest for two primary rea sons: volcanism in the northern portion ofthe arc-backarc region is most likely to have been influenced by the recent rifting of the arc (Stern et al. 1985, Bloomer et al. 1989, Lin et at. 1989); and the northern volcanoes show increasingly alkalic compositions, reaching FIGURE 4 . SeaM AR C II bath ymetry map of Kasuga shoshonitic rock types north of 23° N. The volcano (K I), northern Mariana island are, and two adja cent volcanoes located immediately south of Kasuga, alkalic compositions are suggested to indicate con toured in 100-m intervals.Map shows locations of an incipient rifting environment. Structurally Alvin dives (solid lines with dive numbers) made on these the northern portion of the arc is of great volcanoes . Dives were concentrated on the flanks and interest because the whole of the forearc and summi t regions of the two southern volcanoes in the chain, informally named Kasuga 2 (K2) and Kasuga 3 arc region apparently has been subjected to (K3) . some degree of both cross-arc rifting and forearc extension (Hussong and Fryer 1983, Beal 1987, Fryer et al. 1987a , Mahoney and Fryer 1988). Two large chains of well- Review Article: Marine Geological Research-FRYER 107

develop ed, large volcanic edifices are located tions of the upper half of the deposit indicate in the northern portion of the backarc basin, that it may have been emplaced hot in the extending from the arc line for 60 km into the submarine environment. Large boulders orig backarc basin. The northernmost of these inally up to 2 m in diameter were observed on two, the Kasuga chain, extends onto the rift the surface of the deposit. These had appar axis region (Beal 1987). The Kasuga volcanoes ently decrepitated, and they, as well as were initially studied by SeaMARC II surveys smaller, unbroken boulders, were encrusted and dredging (Hussong and Fryer 1983, Beal on their lower surfaces with hydrothermal de 1987,Jackson 1989). The structural complexi posits . A comparison of the small grain-size ty of the edifices is evident on the side-scan fraction of this deposit with that of Kick-em sonar imagery. There are numerous lava flow Jenny Volcano of the Lesser Antilles arc fields and regions mantled with volcaniclastic shows remarkable similarity (Ballance et al. debris on the two southernmost of the three 1988). volcanoes imaged (Hussong and Fryer 1983). Dredge results show typical backarc basin la SUMISU RIFT, IN THE IZU-BONIN ARC SYSTEM. vas on the lowest slopes of the second volcano The Izu-Bonin backarc region at about 310 N in distance from the active arc (Kasuga 2) (Figure 5) is undergoing the earliest stages of and basalt to andesitic lavas with variable K backarc spreading (Brown and Taylor 1988). content (ranging from < 1% to > 2% with The rift grabens in the area are deeply sedi NazO/KzO < I) (Jackson 1989). Basalts and mented with hemipelagic and volcaniclastic hornblende-bearing dacites were dredged from sediments. However, there are numerous ac the summit region of the third volcano in tive volcanic centers, generally erupted along distance from the arc, Kasuga 3. Magma mix either arc-parallel normal faults or along trans ing processes are evident in the composition of fer zones within the rift basins , and regions of Z the summit lavas from both volcanoes. The high heat flow (up to 600 mWjm ) within the freshness of the lavas recovered and the pres grabens that attest to the recent volcanic ac ence of active hydrothermal systems suggest tivity (Taylor 1987). Several of these neovol that these seamounts are active. The volca canic zones were investigated during an Alvin noes were probably built on backarc crust cruise, augmented by detailed SEABEAM and after a recent cross-arc rifting event, and the bottom camera surveys (Taylor 1987). The magmas from which they were constructed objectives ofthe cruise were to investigate the were probably derived from at least two sepa structure, petrology, and geochemistry of both rate sources, a backarc basin basalt source igneous materials and hydrothermal deposits and an arc source . The Alvin dive cruise to associated with volcanism on the backarc rift these volcanoes and additional bottom pho centers. Previous side-scan sonar studies and tography work and dredging confirm the com dredging of the central rift region showed the plex interrelationships ofmagma sources that Sumisu region at 31005' N to be dominated by had been seen in the dredge samples (Jackson an enechelon ridge with a right lateral offset 1989). Samples collected on the Alvin dive (Brown and Taylor 1988, Brown et al. 1988, cruise further document diversity oflava types Taylor et al. 1988). Most ofthe lavas dredged on the lower flanks of the volcanoes (Jackson from the ridge are backarc basin basalts, simi et al. 1987, Jackson 1989). Hydrothermal sys lar to MORB, but with an overprint ofan arc tems were discovered on the summits of the component (Fryer et al. 1985b, Fryer et al. in two volcanoes on which the dives were con press b). Similar lavas were also recovered ducted (Vonderhaar et al. 1987). A boiling from the base of the fault scarp bounding the system was identified on Kasuga 2. Hydro eastern side of the Sumisu rift. The presence of thermal fluids collected from the Kasuga 2 typical back arc basin basalt so close to the field have an unusually high and as yet enig active arc indicates a separate magma source matic Mg content (McMurtry et al. [in press]). in close proximity to the arc magma source. A large volcaniclastic debris deposit mantles Petrologic studies indicate a shallower source the southern flank of Kasuga 2, and observa- of melting to form the backarc basin basalts of 108 PACIFIC SCIENCE, Volume 44, April 1990

14 0* 00' E 140*20'[

.:\..

~ ..:~., "

FIGURE 5. SeaMARC II bathymetry map of the Sumisu rift contoured in IOO-m intervals, showing locations (boxes) of the regions investigated with Alvin (after Smith 1989). the Sumisu rift (Fryer et al. [in press b]). At the [in press b]): The results of the dive series offset region between the northern and south confirm the previous petrologic findings and ern segments of the ridge, samples of both provide a detailed petrogenetic model based basalt and silicic (dacite and rhyolite) lavas on stratigraphy and clearly defined spatial dis were recovered (Fryer et al. 1985b, Fryer et al. tribution (Hochstaedter et al. [in press]). Hy- Review Article: Marine Geological Research-FRYER 109 drothermally generated barite-silicate chimney an active supra-subduction zone. The interior materials were discovered on the large edifice of Conical Seamount is composed of con centered on the offset region and along the voluted serpentinite flows bearing clasts of southern portion of the northern arm of the metamorphosed ultramafic to mafic rocks. enechelon axial ridge (Urabe and Kusakabe The mafic rocks are of both island arc and [in press]). The hydrothermal deposits are sug midocean ridge compositions. Pore fluids gested to be products of Kuroko-type envi from the serpentine mud flows have composi ronments, and the Sumisu rift represents a tions even more dramatically different from likely site for the in situ study of the kind of that of seawater than those observed seeping high-grade ore generation typical of Kuroko from the chimney structure studied with Alvin deposits. A return to the Sumisu rift region in (Fryer and Pearce et al. 1989a,b). The summit the summer of 1989 with the Shinkai 2000 region of the seamount was confirmed to be submersible (operated by JAMSTEC) pro actively venting fluids and to be producing vided additional evidence for the widespread serpentinite flows through what is interpreted occurrence of barite-silica hydrothermal to be a summit conduit system. Also con chimney deposits in the northernmost part of firmed was the deep-seated origin and the high the southern arm of the Sumisu ridge (A. hydrocarbon content of the associated fluids Malahoff, 1989, pers. comm.). (Fryer and Pearce et al. 1989a,b) . Drilling at the site on the lowermost west flank of Coni cal Seamount intersected a Pleistocene basalt Recent Ocean Drilling Project Studies lava of island arc composition (Fryer and Pearce et al. 1989a,b), thus placing in question Several deep-sea drilling cruises have been the current models of thermal structure ofand conducted in the Philippine Sea area, concen magma genesis in supra-subduction zones. trating on the evolution of the Mariana and Drilling at two lower flank sites on Torishima Izu-Bonin arcs and their backarc basin s. Deep Forearc Seamount, a serpentinite seamount Sea Drilling Project legs 31, 58, 59, and 60 on the Izu-Bonin forearc, revealed serpentinite studied the evolution of the Philippine Sea, flows similar to those on Conical Seamount the Parece Vela Basin, the Shikoku Basin, and beneath the sediment cover. However, in con the Mariana island arc system. Summaries of trast with the active Conical Seamount, the the results of these drilling studies are pre Torishima Forearc Seamount was formed sented by Karig (1975), Klein and Kobayashi some 10 mya and is apparently no longer a (1980), Scott et al. (1980), Natland and Tarney source of serpentinite debris or mud flows (1981), and Hussong and Uyeda (1981). With (Fryer and Pearce et al. 1989a,b). Three sites the exception of leg 60, these early drilling in the Izu-Bonin forearc sediments revealed a cruises focused on the evolution of the backarc history of volcanism from Eocene to Recent basin settings and on the composition of the with peak periods in the Eocene-Oligocene ancient remnant volcanic arcs associated with and Miocene to Pleistocene (Fryer and Pearce them. Leg 60 also investigated the active arc et al. 1989a,b) . The last site drilled penetrated and the forearc of the Mariana system. deep into a Middle Eocene ( ~ 42 mya) volcanic Recently, Ocean Drilling Project legs 125 center, documenting the oldest volcanism and 126 studied the Mariana and Izu-Bonin yet recovered from either the Mariana or Izu forearc regions and the Sumisu rift (Fryer and Bonin forearc. The age of the microfossils Pearce et al. 1989a,b, Fujioka and Taylor et associated with the lavas and the composition al. [in press]). On leg 125 four sites in the of many of the lavas and underlying dikes Mariana forearc and five sites in the Izu (boninitic) are consistent with the formation Bonin forearc provided the first look at the of this volcanic center during the earliest stages interior of forearc serpentinite seamounts, of subduction within the Mariana and Izu allowed documentation of the volcanic base Bonin systems. Thick sequences of pelagic/ ment and its sedimentary history, and re volcaniclastic sediments were drilled in the vealed the first evidence of recent volcanism in forearc on leg 126, documenting Oligocene 110 PACIFIC SCIENCE, Volume 44, April 1990 through Pleistocene volcanism on the arc, and stand the interaction between arc-generated backarc basin sites were drilled revealing rapid lavas and the backarc basin basalts. This in sediment accumulation and the nature of the teraction is a tracer of the tectonic evolution basement volcanics ofthe Sumisu rift (Fujioka of the backarc basin systems. The apparently and Taylor et al. [in press]). shallow-generated backarc basin basalts are emplaced in remarkably close proximity to true arc-generated lavas. Volcanoes of the Summary and Conclusions backarc basins are composed of lavas of a Profound changes in the models ofarc evo wide range of compositions and support a lution have resulted from recent studies ofthe variety of types of hydrothermal systems. Al Mariana and Izu-Bonin interoceanic arc sys though analogies to the kinds ofhydrothermal tems. These arcs are now accepted as having systems found on the midocean ridges can be formed primarily by tectonic erosion, not ac drawn, those of the backarc basin spreading cretion. Their forearc regions are recognized centers have a unique composition reflecting as having migrated north considerable dis influence of the adjacent arc environment. tances since the Eocene. Furthermore, it is ac Even the vent fauna associated with these sys knowledged that the forearcs have experienced tems is in some instances unique. small-scale rotation, vertical deformation, and The importance of understanding these re episodes of rifting in response to plate adjust gions is becoming increasingly evident as more ments , subduction of plate seamounts, and and more ophiolite terrains are identified as collisions of the arc with large aseismic ridges having formed in an arc environment. Serpen on the oceanic plate. Vast portions of the tinite terrains, previously thought to have supra-subduction zone have been subjected to formed during obduction episodes or to be metamorphism driven primarily by fluid flux exposures ofancient transform faults, are now through that region in response to dehydra being recognized as having similarities in struc tion of the subducted oceanic slab. It is likely ture and composition to those described from that the large amounts of fluid hypothesized the Mariana system. Thus, the comparisons of to have been driven otfthe slab over the last 40 subaerial terranes to the recent studies of the million years could be completely accommo interoceanic convergent margins will continue dated by that metamorphism and by the type to add immeasurably to our understanding of of escape of fluids observed at Conical Sea plate interactions and the evolution of the mount. Thus, the problematical imbalance volcanic arcs and backarc basins generated between the volume of fluid derived from the within such environments. slab and the comparatively small amount of volatiles measured in arc-generated lavas can be resolved. The wide variety of rock types LITERATURE CITED recovered from the Mariana and Izu-Bonin forearc and the discovery of ancient volcanic BALLANCE, P. F., S. CAREY, H. SIGURDSSON, centers and recent basalts require us to reeval and P. FRYER. 1988. Production of basaltic uate our concepts of the petrogenetic evolu sediment on Kick 'em Jenny submarine vol tion ofthe forearc crust and mantle. The pres cano, Lesser Antilles arc, and Central Sea ence ofmetamorphosed mantle rocks in great mount, Mariana arc. Abstracts with pro abundance in the outer portions of the fore gram, IGC. arcs will only be fully understood with more BEAL, K. L. 1987. The opening of a backarc detailed study of their composition, distribu basin: Northern Mariana Trough. M.S. tion , and relationships to structural features thesis, University of Hawaii. of the forearcs, BECCALUVA, L., G. MACCIOTrA, C. SAVELL!, The study of the arc lavas of the northern G. SERRI, and O. ZEDA. 1980. Geochemistry Mariana arc, the Mariana backarc basin, and and K/Ar ages of volcanics dredged in the the Sumisu rift in the Izu-Bonin backarc re Philippine Sea (Mariana, Yap, Palau trench gion provides the means by which to under- es and Parece-Vela Basin). Pages 247-270 Review Article: Marine Geological Research- FRYER III

in D. E. Hayes, ed. The tectonic and geo Trough: Results of the first Alvin dives . logic evo lution ofSoutheast Asian seas and Eo s 68(44): 1531. islands. Geophys. Monogr. Ser., Vol. 23. DADEY, K . A., and M . LEINEN. 1987. Anoma AGU, Washington, D .C. lous porosity in Mariana mounds sedi BLOOMER, S. 1982. Structure and evo lution of ment s. Eos 68(44): 1534. the Mariana Trench: Petrologic and geo D IETRICH, V.R., R. EMMERMAN, R. OBER chemical studies. Ph.D. dissertation, Uni HANSLI, and H. PUCHLET. 1978. Geochem versity of California, San Diego. istry of basaltic and gabbroic rocks from BLOOMER, S. H ., and J. W. HAWKINS. 1983. the west Mariana Basin and the Mariana Gabbroic and ultramafic rocks from the Trench. Earth Planet. Sci. Lett. 39: 127- 144. Mariana Trench: An island arc ophiolite. FRYER, P., and G . J. FRYER . 1987. Origins of Pages 294-317 in D . E. Hayes, ed. The tec nonvolcanic seamounts in a forearc envi tonic and geologic evo lution of Southeast ronment. Pages 61-72 in B. H . Keating, P. Asian seas and islands, Part 2. Geophys. Fryer, R . Batiza, and G . W. Boehlert, eds . Monogr. Ser ., Vol. 27. AGU, Washington, Seamounts, islands and atolls. Geophys. D.C. Monogr. Ser., Vol. 43. AGU, Washington, ---. 1987. Petrology and geochemistry of D .C. boninite series volcanic rocks from the FRYER, P., and D . M . HUSSONG . 1985. Sea Mariana trench. Contrib. Mineral. Petrol. MARC II studies ofsubducting seamounts. 97 :361 -377. Pages 291- 306 in N . Nasu, ed. Formation BLOOMER , S. H. , R.J. STERN, E. FISK , and of active ocean margins. Terra Scientific C. H . GESHWIND. 1989. Sho shonitic volcan Pub!. Co ., Tokyo. ism in the northern Mariana Arc. 1. Min FRYER, P., and N. C. SMOOT. 1985. Morpholo era logic and major and trace eleme nt gy ofocean plate seamounts in the Mariana characteristics. J. Geophys. Res . 94(B 4): and Izu -/Bonin subduction zone. Mar. 4469-4496. Geo!' 64 : 77- 94. BROWN, G ., and B. TAYLOR. 1988. Seafloor FRYER, P., E. L. AMBOS, and D . M . HUSSONG. mapping ofthe Sumisu Rift, Izu-Ogasawara 1985a. Origin and emplacement ofMariana (Bonin) island arc.Bull. Geol. Surv. Jpn. forearc seamounts. Geology 13:774-777. 39: 23-38. FRYER, P., J. GILL, M . JACKSON, R. FISKE, BROWN, G ., ET AL. 1988. Geological map A. HOCHSTAEDTER, G. McMURTRY, P. of Sum isu and Torishima Rifts, Izu SEDWICK, A. MALAHOFF, P. MOUGINIS Ogasawara Arc. Marine Geological Map MARK, and E. HORIKOSHI. 1987a. Results of Series 31. Geological Survey ofJapan. recent Alvin studies of the Kasuga volca CAMPBELL, A. c., J. M . EDMOND, D. noes , northern Mariana arc. Eos 68(44): COLODNER, M . R .PALMER, and K . K . 1531. FALKNER. 1987. Chemistry of hydrother FRYER, P., J. A. HAGGERTY, B. TILBROOK, ma l fluids from the Mariana Trou gh P. SEDWICK, L. E. JOHNSON, K . L. SABODA, backarc basin in comparison to mid-ocean S. Y. NEWSOM, D. E. KARIG, S. UYEDA, and ridge fluids . Eos 68(44): 1531. T.ISlliI. 1987b. Results ofstudies ofMariana CARLSON, C. 1984. Stratigraphic and structur forearc serpentinite diapirism. Eos 68(44): al significance offoliate serpentine breccias. 1534. Pages 108-112 in Field Trip Guidebook, FRYER , P., J. HAGGERTY, B. TILBROOK, P. no . 3. Society of Economic Paleontologists SEDWICK, K . L. SABODA, G. M cMURTRY, and Mineralogists, Wilbur Springs. L. E. JOHNSON, S. Y. N EWSOM, D .E. KARIG, CLOUTIER , M .J., and J. A. HAGGERTY. 1988. S. UYEDA, T. ISHII, and D . STAKES. Alvi n Bacteria associated with Mariana forearc observations of serpentinite seamounts on chimneys. Abst. 88th Annu. Meet. Am . the Mariana convergent margin. Earth Soc. Microbiol. 88: 248. Planet. Sci. Lett. (in press a). CRAIG, H ., Y. HORIBE, and K . A. FARLEY. FRYER, P., C. LANGMUIR, B. TAYLOR, Y. 1987. Hydrothermal vents in the Mariana ZHANG, and D . M. HUSSONG. 1985b. Rift- 112 PACIFIC SCIENCE, Volume 44, April 1990

ing of the Izu arc, Ill: Relationship of munities ofthe Mariana backarc basin. Eos chemistry to tectonics. Eos 66(18): 421. 68(44): 1531. FRYER, P., and J. A. PEARCE and THE ODP HOBART, M. A., R . N . ANDERSON, N . FUJII, LEG 125 SCIENTIFIC SHIPBOARD PARTY. and S. UYEDA. 1983. Heat flow from hydro 1989a. Ocean Drilling Program plumbing thermal mounds in two million year old the Pacific sinks. Nature (London) 339: crust ofthe Mariana Trough which exceeds 427-428. two watts per meter. Eos 64: 315. - - . 1989b. In the Pacific ODP Leg 125 HOCHSTAEDTER, A. G ., J. B. GILL, C. drills forearc crust, mantle. Geotimes 34(7): LANGMUIR, P. FRYER, and B.TAYLOR. 1987. 18-20. Alvin studies of the volcanic geology and FRYER, P., J. M. SINTON, and J. A. PHILPOTTS. petrology of the Sumisu rift. Eos 68(44 ): 1981. Basaltic glasses from the Mariana 1534. Trough. Pages 601-607 in D. M. Hussong HOCHSTAEDTER, A. G., J. B. GILL, C. H. and S. Uyeda, eds. Initial reports of the LANGMUIR, J. D. MORRIS, and M. PRINGLE. Deep-Sea Drilling Project, Vol. 60. U.S. Petrology and geochemistry of lavas from Govt. Printing Off., Washington, D.C. the Sumisu rift: An incipient backarc basin. FRYER, P., B. TAYLOR, A. G . HOCHSTAEDTER, Earth Planet. Sci. Lett. (in press). and C. LANGMUIR . Petrology and geochem HONZA, E., and K: TAMAKI. 1985. The Bonin istry of lavas from the Sumisu and Tori arc . Pages 459-502 in A. E. M . Nairn and S. shima backarc rifts . Earth Planet. Sci. Lett. Uyeda, eds. The ocean basins and margins. (in press b). Vol. 7. The Pacific Ocean. Plenum Press, FUJIOKA, and B. TAYLOR and THE ODP LEG New York. 126 SCIENTIFIC SHIPBOARD PARTY. Arc vol HORIBE, Y , K. KIM, and H. CRAIG. 1986. canism and rifting in the Izu-Bonin forearc Hydrothermal methane plumes in the and backarc. Nature (London) 342: 18-20. Mariana backarc spreading center. Nature HAGGERTY, J. A. 1987a. Petrology and geo (London) 324: 131-133. chemistry of neogene sedimentary rocks HUSSONG, D. M ., and P. FRYER. 1983. Back from Mariana forearc seamounts. Pages arc seamounts and the SeaMARC II sea 175-186 in B. H . Keating, P. Fryer, R. floor mapping system. Eos 64(45): 627 Batiza, and G . W. Boehlert, eds. Seamounts, 632. islands and atolls. Geophys. Monogr. Ser., - -- . 1985. Forearc tectonics in the north Vol. 43. AGU, Washington, D.C. ern Mariana arc. Pages 273-290 in N. - -- . 1987b. Cold-water, deep-sea chim Nasu, ed. Formation of active ocean mar neys from the Mariana forearc serpentinite gins. Terra Scientific Publ. Co. , Tokyo. seamounts. Eos 68(44): 1534. HUSSONG, D. M., and S. UYEDA. 1981. Tec - --. 1989. Fluid inclusion studies ofchim tonic processes and the history of the neys associated with serpentinite sea Mariana arc: A synthesis of the results of mounts in the Mariana forearc. Pan Ameri Deep Sea Drilling Project Leg 60. Pages can Conference: Research on Fluid Inclu 909-929 in D . M. Hussong and S. Uyeda, sions (PACROFI) II, Blacksburg, Va. 2:29. eds. Initial reports ofthe Deep-Sea Drilling HAGGERTY, J., and M. J. CLOUTIER. 1988. Bio Project, Vol. 60. U .S. Govt. Printing Off., logic and mineralogic studies of deep-sea Washington, D.C. chimneys venting cold water from Mariana ISHII, T. 1985. Dredged samples from the forearc seamounts. Geol. Soc . Am. Ab Ogasawara forearc seamount or "Paleo stracts with program, 20(7): A 199. land"- "Forearc ophiolite." Pages 307 HAWKINS, J. , J. MELCHIOR, and P. LONSDALE. 342 in N . Nasu, ed. Formation of active 1987. Petrology of axial region of the ocean margins. Terra Scientific Publ. Co ., Mariana Trough. Eos 68(44): 1530-1531. Tokyo. HESSLER, R. R. , S. C. FRANCE, and M . A. JACKSON, M. C. 1989. Petrology and pet BOUDRIAS. 1987. Hydrothermal vent com- rogenesis of recent submarine volcanoes Review Article:Marine GeologicalResearch-FRYER 113

from the northern Mariana arc and back LIN, P.-N., R. J. STERN, and S.H. BLOOMER. arc basin. Ph.D . dissertation, University of 1989. Shoshonitic volcanism in the north Hawaii. ern Mariana arc 2: Large-ion lithophile and JACKSON, M . c., P.FRYER, and J. GILL. 1987. rare earth element abunda nces evidence for Petrology of volcanic rocks collected on a the source of incompatible element enrich recent Alvin cruise to the Kasuga volcanoes, ments in intraoceanic arcs . J. Geophys. northern Mariana arc . Eos 68(44) : 1532. Res. 94(B4) : 4497-4514. JOHNSON, L. E., and P. FRYER. 1988. Oceanic LOCKWOOD, J. P. 1971. Sedimentary and grav plate material on the Mariana forearc. Eos ity slide emplacement ofserpentinite. Geol. 69(44): 1471 . Soc. Am. Bull. 82: 919-936. - -- . Petrography, geochemistry and pe LONSDALE, P., and J. W. HAWKINS. 1985. trogenesis of igneous rocks from the outer Silicic volcanism at an off-axis geothermal Mariana forearc. Earth Planet. Sci. Lett. (in field in the Mariana Trough backarc basin. press). Geol. Soc. Am . Bull. 96 :940-951. KARIG, D. E. 1971. Structural history of the ---. 1987. The geologic setting of hydro Mariana island arc system . Geol. Soc. thermal vents at the Mariana Trough Am. Bull. 83 :323-344. spreading center. Eos 68(44) : 1530. ---. 1975. Basin genesis in the Philippine MACDOUGAL, J.D ., A. VOLPE, and J. W. Sea. in D. E. Karig, ed. Initial reports ofthe HAWKI NS . 1987. An arc-like component in Deep-Sea Dri lling Project, Vol. 3L U.S. Mariana Trough lavas . Eos 68(44): 1531. Govt. Printing Off., Washington, D .C. MAHONEY, J., and P. FRYER. 1988. Stress field KARIG, D. E., and B. RANKEN. 1983. Marine re-orientation of the inner Mariana forearc geology of the forearc region, southern at 22° N . Eos 69(44): 1443. Mariana island arc. Pages 266-280 in D . E. McMuRTRY, G., P. FRYER , and D . Hayes, ed. The tectonic and geologic evolu VONDERHAAR. High and low temperature tion of Southeast Asian seas and islands, hydrothermal venting and deposition at Part 2. Geophys. Monogr. Ser., Vol. 27. Fukujin Seamount, and the Kasuga Sea AGU, Wa shington, D.C. mounts ofthe northern Mariana Arc . Earth KASTN ER, M., H . CRAIG, and A. STURZ. 1987. Planet. Sci. Lett. (in press) . Hydrothermal deposits in the Mariana MEIJER, A. 1980. Primitive arc volcanism Trough: Pre liminary mineralogical investi and a boninite series; examples from west gation. Eos 68(44): 1531. ern Pacific island arcs. Pages 271-282 in D . KLEIN, G. deV. , and K. KOBAYASHI, eds. 1980. E. Hayes, ed. The tectonic and geologic Initi al reports of the Deep-Sea Drilling evolution of Southeast Asian seas and Project, Vol. 58. U.S. Govt. Printing Off., islands. Geophys. Monogr. Ser., Vol. 23. Washington, D.C. AGU, Was hington, D.C. LEINEN, M., and R. N. ANDERSON. 1981. Hy MEIJER, A., E. ANTHONY, and M. REAGAN. drothermal sediment from the Mariana 1981. Petrology of volcanic rocks from the Trough. Eos 62:914. fore-arc sites. Pages 709-729 in D .M. LEINEN, M., R. McDuFF, J. DELANEY, K. Hussong and S. Uyeda , eds. Initial reports BECKER, and P. SCHULTHEISS. 1987. Off-axis of the Deep-Sea Drilling Project, Vol. 60. hydrothermal activity in the Mariana U.S. Govt. Printing Off., Washington, D .C. mounds field. Eos 68(44): 1531. N ATLAND, J. H. , and J. TARNEY. 1981. Petro LEWIS, S. D., D. E. HAYES, and C. L. logic evolution of the Mariana arc and MROZOWSKI. 1982. The origin of the West backarc basin system -A synthesis ofdri ll Philippine Basin by interarc spreading. ing results in the south Philippine Sea. Page-s 31-51 iii G. R. Balce arid A. S: Pages 877-907 in D. M . Hussong and S. Zanoria, eds. Geology and tectonics of Uyed a, eds. Initial reports of the Deep-Sea the Luzon-Marianas region. Philippine Drilling Project, Vol. 60. U .S. Govt. Print SEATAR committee Spec. Publ. 1. ing Off., Washington, D .C. 114 PACIFIC SCIENCE, Volume 44, April 1990

NEWSOM, S. Y., and P. FRYER. 1987. Three TAYLOR, B., G . BROWN, P. FRYER , J. GILL, F. dimensional gravity modeling of serpen HOCHSTAEDTER, H . HOTTA, C. LANGMUIR, tinite seamounts in the Mariana forearc. M. LEINEN, A. NISHIMURA, and T. URABE. Eo s 68(44) : 1534. Alvin-SeaBeam studies of the Sumisu Rift, SABODA, K. L., P. FRYER , and G . FRYER . 1987. Izu-Bonin arc. Earth Planet. Sci. Lett. (in Preliminary studies of metamorphic rocks press). collected during Alvin studies of Mariana URABE, T., and M . KUSAKABE. Barite-silica fore arc seamounts. Eos 68(44): 1534. chimney in the Sumisu Rift, Bonin (Izu SCOTT, R. B., L. W. KROENKE, G. ZAK A Ogasawara) arc; product of Kuroko-type RIADZE, and THE LEG 59 SCIENTIFIC SHIP hydrothermal acti vity. Earth Planet. Sci. BOARD PARTY. 1980. Evolution ofthe South Lett. (in press). Philippine Sea: Deep Sea D rilling Project UYEDA, S., and Z. BEN-AVRAHAM. 1972. Ori Leg 59 results. Pages 803-815 in R. B. Scott gin and development of the Philippine Sea. and L. W . Kroenke, eds . Initial reports of Nature (London) 240: 176- 178. the Deep-Sea Drilling Project, Vol. 59. U .S. VONDERHAAR, D . L., G . M . M cMURTRY, P. Govt. Printing Off., Washington, D .C. FRYER , and A. MALAHOFF. 1987. Geochem SHIH, T.-C. 1980. Marine magnetic anomalies istry of hydrothermal deposits from active from the western Phi lippine Sea. Pages 49 submarine volcanoes in the Mariana arc. 76 in D . E. Hayes, ed. The tectonic and Eo s 68(44): 1533. geologic evolution of Southeast Asian seas WHEAT, C. G. , and R. E. McDUFF. 1987. and islands. Geophys. Monogr. Ser. , Vol. Ad vection of pore waters in the Marianas 22. AGU, Washington, D.C. mounds hydrothermal region as deter SINTON, J. M ., and P. FRYER. 1987. Mariana mined from nutrient profiles. Eos 68(44): Trough lavas from 18° N : Implications for 1531. the origin of backarc basin basalts. J. WOOD , D . A., N . G. MARSH, J. TARNEY, J.-L. Geophys. Res . 92(12): 12,782- 12,802. JORON, P. FRYER , and M . TREUIL. 1981. SKORNYAKOVA , N . S., and M . 1. LIPKINA. Geochemistry of igneous rocks recovered 1976. Basic and ultrabasic rocks of the from a transect across the Mariana Trough, Marianas Trench. Oceanology 15:688-690. arc, forearc, and trench, Sites 453 through SMITH, J. R., Jr. 1989. Geomorphology and 46 1, Deep Sea D rilling Project Leg 60.Pages growth processes ofsubmarine volcanic ter 611-645 in D . M . Hussong and S. Uyeda, rain in a rifting environment: The Bonin eds . Initial reports ofthe Deep-Sea D rilling backarc. M.S. thesis , University ofHawaii. Project, Vol. 60. U .S. Govt. Printing Off., STERN, R . J., N . C. SMOOT, and M . RUBIN . Washington, D .C. 1985. Unzipping ofthe Volcano arc, Japan. YAMANO, M. , N . FUJII, S. NAGIHARA, and M . Tectonophysics 102: 153- 174. LEINEN. 1987. Temperature monitoring in TAYLOR, B. 1987. Active rifting of the Bonin marine sediments in the Mariana Trough. arc. Eo s 68(44): 1534. Eos 68(44): 1534. TAYLOR , B., ET AL. 1988. SeaMARC II ba YAMAZAKI , T ., and Y. OKAMURA . 1989. Sub th ymetry and side-scan acoustic imagery ducting seamounts and deform ation of of Sumi su and Torishima Rifts, Izu overriding forearc wedges around Japan. Ogasawara Arc. Marine Geological Map Tectonophysics 160:207-229. Series 31. Geological Survey ofJapan .