Seafloor Samples Research Articles

Hydrothermal convection at mid‐ocean ridges: Influence of permeability anisotropy within a tilted triangular structure

Recent results from seafloor exploration and crustal drilling are used to constrain the structures within which hydrothermal circulation occurs at mid‐ocean ridges (MOR). This hydrothermal activity is modeled numerically using a two‐dimensional triangular domain representing the extrusive layers. The configuration is heated both at an impermeable base and along the intrusion axis. The structure is characterized by dips of a few degrees towards the axis and permeability anisotropy (δ = KZ/KX). The temperature fields are computed for different hydrothermal regimes defined by variable axial heat flow values characterized by variable Rayleigh numbers (Ra*). These computed temperatures are compared to those deduced from basalt alteration mineralogy observed in seafloor samples and drillcores. The mean temperature of discharge zones is considered as limited to values between 60° and 100°C.In the isotropic case, the chosen geometry and boundary conditions are found to stabilize the convective process, which advocates against a time‐fluctuating behavior as a steady state is reached in an average time of 5000 years up to Ra* values as large as 600. However, Ra* values are unlikely to exceed 400, or to be smaller than 100 due to the mineralogical constraints. With Ra* larger than 100 and δ=1, a mean temperature of 25°C, compatible with celadonite‐nontronite and iron‐hydroxides mineralization, is obtained for the recharge zones. For a given axial heat flow and a similar permeability contrast, larger temperatures are obtained in subhorizontal settings than in subvertical ones. Small changes of permeability anisotropy are confirmed to induce large variations in temperatures and strength of the flow.

High Resolution Surficial Geology of the Louisiana Middle to Upper Continental Slope: ABSTRACT

ABSTRACT Newly acquired high resolution acoustic data coupled with direct sea floor observational data and sampling by research submersibles have provided a new appreciation of the local-to-mesoscale geologic complexity of the ocean floor in middle and upper continental slope depths (depths 200-1000m). This relatively small-scale geologic complexity is superimposed on regional geologic features dominated by salt diapirs and adjacent intraslope salt/shale withdrawal basins. While the sea floor of intraslope basins is largely characterized by featureless depositional topography, surficial geology is extremely variable on the tops and flanks of salt diapirs. Dome tops are eroded and commonly covered with a coarse sediment lag characterized by both biogenic carbonates and clasts of diagenetic origin. Intense faulting associated with diapirs exposes uplifted, tilted, fractured, and sometimes folded slope sediments. These faults provide avenues for the transport of fine-grained sediments, formation fluids, brines, and hydrocarbons to the sea floor. Depending on transported products and flux rate, sea floor features range from massive mud volcanoes (> 1km diameter) to carbonate mounds ( 10m relief) and hardgrounds (byproducts of microbial degradation of hydrocarbons). On the uppermost slope, salt dome tops are sites of biogenic carbonate buildups and carbonate veneering formed during periods of lowered sea level during the late Pleistocene. Below water depths of about 500m moderate-to-slow vertical flux of hydrocarbons frequently results in the formation of gas hydrates at or near the sea floor. Hydrate mounds and families of mounds possess surface features that reflect brine and fluid mud expulsion, authigenic carbonate formations, and benthic communities that depend on chemical constituents of the hydrate complex. Local over-steepening of diapirs, hydrate mounds, and large mud vents causes slope instability resulting in a wide range of mass movement features from those associated with surface creep to huge submarine landslides.

Correction [to “Characterizing potential offshore seismic sources using high‐resolution geophysical and seafloor sampling programs: An example from Cape Egmont fault zone, Taranaki shelf, New Zealand” by Scott D. Nodder

Correction [to “Characterizing potential offshore seismic sources using high‐resolution geophysical and seafloor sampling programs: An example from Cape Egmont fault zone, Taranaki shelf, New Zealand” by Scott D. Nodder

Characterizing potential offshore seismic sources using high‐resolution geophysical and seafloor sampling programs: An example from Cape Egmont fault zone, Taranaki shelf, New Zealand

Standard marine geological techniques (high‐resolution seismic reflection methods, piston cores) are used to investigate the neotectonics and paleoseismology of an active part of an offshore fault system, the Cape Egmont fault zone, Taranaki continental shelf, New Zealand. The main active structure within the zone, the Cape Egmont fault, is a reactivated Pliocene to Recent normal to oblique‐slip fault that is associated with shallow crustal seismicity and exhibits a 53‐km long, 1‐to 5‐m high, low‐angle drape fold or “scarp” at the seafloor. Four, possibly five, fault segments varying in length from 7–51 km are delineated using geometric, structural, and stratigraphic information. Empirical estimates of maximum credible earthquake magnitudes, based upon seismic moment calculations, surface rupture lengths, and vertical surface offsets, suggest that the prehistoric earthquake(s) that formed the present seafloor surface trace probably had a magnitude of at least MW 6.4(MW, moment magnitude) and could have been as large as MS 7.3–7.4(MS, surface wave magnitude). Tentative maximum recurrence intervals for such events, inferred for each segment, range from 8000–13,000 years, with lower magnitude events having correspondingly shorter recurrence intervals of 500–900 years. Without careful study of offshore late Quaternary faults and fault zones, particularly where they are contiguous with areas of active onshore deformation, existing regional seismic hazard predictions may be inaccurate.

Very High-Resolution Seismics as a Tool for Sequence Stratigraphy Applied to Outcrop Scale--Examples from Eastern Tyrrhenian Margin Holocene/Pleistocene Deposits

Three Eastern Tyrrhenian continental margin areas with very different depositional and tectonic settings were analyzed using single-channel very high-resolution reflection seismics coupled with sea-floor sampling. Fourth- and fifth-order depositional sequences derived from Quaternary glacio-eustatic processes are now buried at shallow depths within the continental shelf. The sequences have been reconstructed in detail with great lateral continuity. The tectonic setting did not influence the development of high-order sequences, which are probably synchronous at a global scale. Within these sequences, it has been possible to recognize all the systems tracts hypothesized by sequence stratigraphy models, confirming that this method does not depend upon time or scale. The geom try of depositional sequences is strongly dependent on sedimentation rate, i.e., on the distance from sources. Local factors such as transport pattern variations caused by sea level changes, paleotopography of the transgressed surface, characteristics of feeding hydrographic nets, and interactions with bioconstructions must be considered because they play an important role in the geometrical development of depositional sequences. In high-order depositional sequences, lowstand systems tracts are volumetrically predominant. Highstand and transgressive systems tracts seem to be linked to point sources because they have a three-dimensional (3-D) geometry. On the contrary, lowstand systems tracts are related to linear sources parallel to the shelf break; they cause the progradation of the whole margin so that they form the margin structure almost entirely.

Porosity determinations of sediments in Knight Inlet using a transient electromagnetic system

A newly developed towed transient electromagnetic (TEM) system is capable of measuring the electrical conductivity of the uppermost 5–10 m of sediments on the sea floor. The profiles of conductivity may be interpreted to give the porosity and likely texture of the bottom sediments. Recent tests of the system have demonstrated that data may be collected continuously in a surveying mode. Results from Knight Inlet, British Columbia, are in good agreement with sea floor samples from the area.

Next generation of seafloor samplers

The University of Washington, Seattle, has successfully completed the first stage of a proposed multiphase program to develop small rock drills as a new generation of seafloor sampling tools. Phase I of this program, a basic 3‐m drill, was funded by the National Science Foundation, the Office of Naval Research, and Washington Sea Grant. The drill was designed and built for the university by Williamson and Associates of Seattle and was successfully deployed, in 2400m of water on the Juan de Fuca Ridge, during August 1990. The present drill can penetrate and sample 3 m of vertically oriented core from crystalline igneous rocks, lithified sediments, and sulfide and carbonate deposits. It is deployable to a depth of 5000 m (limited by the cable and drill weight) by conventional research vessels using the NSF‐standard armored conducting cable. The drill is presently undergoing post‐test modifications at the University of Washington and will be available to the scientific community through a Regional Facility by summer 1991.

Side-Scan Sonar and Submersible Observations: New Techniques for Gleaning More Information from Sea-Floor Outcrops: ABSTRACT

Advances in high resolution side-scan sonar imaging technology are so effective at imaging sea-floor geology that they have greatly improved the efficiency of a bottom sampling program The traditional sea-floor geology methodology of shooting a high-resolution seismic survey and sampling along the seismic grid was considered successful if outcrops were sampled on 20% of the attempts. A submersible was used sparingly because of the inability to consistently locate sea-floor outcrops. Side-scan sonar images have increased the sampling success ratio to 70-95% and allow the cost-effective use of a submersible even in areas of sparse sea-floor outcrops. In offshore basins this new technology has been used in consolidated and semiconsolidated rock terranes. When combined with observations from a two-man submersible, SCUBA traverses, seismic data, and traditional sea-floor bottom sampling techniques, enough data are provided to develop an integrated sea-floor geologic interpretation. On individual prospects, side-scan sonar has aided the establishment of critical dip in poor seismic data areas, located seeps and tar mounds, and determined erosional breaching of a prospect. Over a mature producing field, side-scan sonar has influenced the search for field extension by documenting the orientation and location of critical trapping cross faults. These relatively inexpensive techniques can providemore » critical data in any marine basin where rocks crop out on the sea floor.« less

The flux and preservation of organic carbon in Baffin Island fjords

Abstract Except for shell hash lags found along minor portions of fjord walls and on current swept sills, organic carbon dominates over inorganic carbon, accounting for between 95 and 1000f the total carbon composition within the fjords. The deposition of terrestrial organic carbon is of minor significance except proximal to river mouths; i.e. the atomic carbon/nitrogen ratio, both within the water column and within the fjord sediments, fluctuates around 6.4. Organic carbon is an important constituent of the suspended particulate matter throughout most of the year when the suspended sediment concentration is < 2 g m −3 . During the summer meltwater season, the organic fraction of the suspended sediment is overwhelmed by inorganic mineral matter. In the autumn season, the organic carbon content of the suspended particulate matter can exceed 30%, especially in the seasonal layer where productivity is highest. Bacterial biomass accounts for more than 500f the total biomass found at the sediment-water interface. Microbial respiration of organic carbon is highly effective, consuming > 900f the annual deposition of carbon to the sea floor. The remaining and mostly refractory carbon preserved within the sediment column is < 20y weight of the sediment and, in high sedimentation environments, such as proximal to tidewater glaciers, organic carbon is < 0.1%. The organic carbon content of sea floor samples is found inversely proportional to the modern suspended load delivered to these fjords. Anoxic degradation of this buried carbon appears to be a slow and ineffective process over the time period of 10 000 years. The abundance of organic carbon buried within Arctic fjord sediments is therefore considered as an inverse indicator of deglacial and post-glacial sedimentation rates. A consistent organic carbon stratigraphy relating to the deglacial and Hypsithermal period, the Neoglacial period culminating in the Little Ice Age, and the present-day warm conditions has been identified in all ten Baffin Island fjords investigated. A simple technique is presented to convert the content of organic carbon observed down core to relative or absolute sedimentation rates, depending on the availability of 14 C dates. Instantaneous deposition of sediment from slides or gravity flows remains problematic.

Seismic reflection characteristics of glacial and glacimarine sediment in the Gulf of Alaska and adjacent fjords

Glaciation together with tectonism have been dominant factors affecting sedimentation in the Gulf of Alaska area from at least the late Miocene throughout the Quaternary. The effects of tectonism are apparent in high mountains that border the gulf, raised terraces of Middleton Island and the eastern gulf coastal zone, and numerous active faults and related earthquakes. Glacial evidence includes magnificent glaciers and their onshore deposits, spectacular fjords, large sea valleys incised in the continental shelf, submarine morainal ridges at mouths of bays and sea valleys, and thick glacimarine sedimentary sequences (diamicts) that are exposed onshore and at the sea floor along the outer shelf. Seismic-reflection profiling and sampling of the uppermost marine sedimentary sequences in the Gulf of Alaska and adjacent fjords and bays have allowed identification of three discrete glacially related stratigraphic units. These units were delineated on the basis of seismic signature, geometry, physiographic location, stratigraphic position, and sedimentologic characteristics. The oldest unit, a Quaternary diamict, is portrayed on seismic profiles by irregular, discontinuous reflections. This unit probably includes till, outwash and glacimarine sediment. A geographically restricted unit, one incorporating Holocene end moraines at bay mouths and associated with some sea valleys, consists of jumbled masses of discontinuous reflections and very irregular surface morphology. The youngest unit, a blanket of Holocene sand to clayey silt prograding as a sediment wedge across the shelf, contains nearly horizontal, parallel reflections except where disrupted by mass movement. Although seismic-reflection data alone cannot provide definitive proof of the presence of glacial sediment, when combined with sea-floor sampling, seismic profiling is a powerful tool for determining the continuity of marine sedimentary units and relationships to past and modern glaciers.

C 1[sbnd]C 8 hydrocarbons in sediments from Guaymas Basin, Gulf of California—Comparison to Peru Margin, Japan Trench and California Borderlands

Surface seafloor sediments, hydrothermal vent samples, and Deep Sea Drilling Project sediments (Hole 481A) from the Guaymas Basin were examined for C 1 C 8 hydrocarbons. The proportions of various classes of compounds were examined and compared to those from other geographic areas (Peru upwelling region and Japan Trench) to gain insight into the relative importance of thermal generation, migration and biodegradation. Concentrations of C 2 C 7 hydrocarbons were about 10–10,000 times higher in geothermally warm (estimated to have been exposed to maximum temperatures in the range of 30–150°C) Guaymas Basin sediments in comparison to the low concentrations (0.1–10 ppb per compound) typical of geothermally cold (maximum thermal exposure less than 20°C) seafloor and DSDP diatomaceous sediments. However, one sediment sample from DSDP Site 477, estimated to have been exposed to temperatures of 300°C or higher in the past, showed only a limited hydrocarbon composition, consisting of C 1 C 3 alkanes and aromatic hydrocarbons only. Alkene/alkane ratios of 0.1 or greater were typical of both geothermally cold sediments and also of very hydrocarbon-rich Alvin samples recovered from the seafloor. Because little or no alkene was generally detected in buried sediments exposed to geothermal temperatures greater than 30°C, it is suggested that the alkenes are produced by biogenic processes. Normal alkenes predominated over cyclic and branched structures in geothermally cooler (<20°C) sediments, with the proportion of cyclic and branched compounds increasing in hotter sediments. Concentrations of gem-dimethyl and aromatic compounds generally remained approximately constant or increased slightly with temperature in comparison with geothermally cold shallow sediments. Similarities in compositions of branched and cyclic compounds were observed in some pairs of bitumen-rich Guaymas seafloor samples recovered from different areas, suggesting common mechanisms of light hydrocarbon generation and/or migration. Localized increases in ratios of specific cycloalkane ratios were observed adjacent to sill intrusions.

Plankton percentage of the foraminiferal fauna in seafloor samples from the Australian-Irian Jaya continental margin, eastern Indonesia

Plankton percentages of the formainiferal fauna in seafloor samples from the Australian-Irian Jaya continental margin show a consistent relationship with waterdepth. Based on this relationship a regional model has been constructed to estimate paleowaterdepth from the plankton percentage of fossil foraminiferal fauna. In addition to paleobathymetric information obtained from the study of benthic foraminifera, this model can be applied to reconstruct the paleoenvironment of Late Cenozoic sections sampled on the eastern Indonesian islands.

Influence of bottom currents on sediment texture and sea-floor morphology in the Argentine Basin

Summary The relative bottom-current speed below 4000 m in the Argentine Basin was inferred from the regional pattern of mean particle size of the non-biogenic silt fraction of about 300 sea-floor samples. The pattern reveals a strong deep western boundary current (DWBC) entering the Argentine Basin from the Georgia Basin through a gap in the Falkland Fracture Zone. The DWBC turns W and flows as a contour current along the Falkland Escarpment and Argentine continental margin. At the southern end of the Rio Grande Rise, the DWBC is deflected to the E and SE where it flows along the lower flank of the Mid-Atlantic Ridge as a weak basin-wide return flow. Several smaller return-flow gyres in the southern and western basin are delineated in the regional pattern. The DWBC is strongest along the western margin of the Argentine Basin where 3.5 kHz echograms reveal very prolonged echoes with no sub-bottom reflectors. That reflector pattern is indicative of coarse turbidites, some of which have been winnowed to produce coarse-grained lag deposits under the axis of flow. In the interior of the basin, where bottom-current flow is weakest, sediment consists of clay and silt which is deposited in large migrating mud waves with wavelengths of 3–10 km and heights of up to 137 m (average 26 m). The mud waves consist of material swept into the basin by the DWBC and delivered to the basin margin by down-slope processes. Thus material is winnowed by the DWBC and deposited as a fine-grained chaff in abyssal antidunes which migrate to the centre of the basin and form thick drift deposits.

Introduction to Special Section on Open Magmatic Systems

The idea that magmatic systems are open to intermittent gains and losses of mass and energy is hardly new. Eruptive, conductive, and hydrothermally convective losses are obvious, and both wall rock assimilation and mixing of discrete magma batches are petrological ideas with long and distinguished conceptual lineages. New ideas of the last decade are largely outgrowths of the mounting evidence that mixing and assimilation are ubiquitous, that heterogeneity and mixing in mantle source regions are common, and that few igneous rock suites are likely to reflect strictly closed‐system fractionation of single magma batches.The evidence has accumulated quickly on many fronts, owing in part to the wider availability of high‐precision analytical tools and the development of sophisticated methods for modeling the wealth of chemical and isotopic data. Detailed studies of zoned ash flow sheets, zoned and layered plutons, and macroscopically mixed igneous rocks have made petrologists aware that dynamic and nonequilibrium aspects of magma transport play important roles in the evolution of many magmatic systems. Technological advances have permitted routine experimentation over a wide range of pressures, providing kinetic data and phase equilibrum constraints essential to understanding both mantle source processes and magmatic evolution in crustal reservoirs. Finally, the plate tectonic synthesis and seafloor sampling programs have respectively provided the impetus for a detailed assessment of geochemical heterogeneity in a dynamic mantle and an important avenue for accomplishing that assessment.

A Late Pleistocene time-series of bottom-current speed in the Vema Channel

A coarsening of the mean particle size of the carbonate-free silt fraction from sea-floor samples below 4000 m in the Vema Channel has been used to separate high-velocity Antarctic Bottom Water (AABW) from the overlying, slower North Atlantic Deep Water (NADW). A time-series of fluctuations in bottom-current speed within the modern AABW/NADW transition zone was examined by determining the particle-size distribution of sediments from eight gravity cores with a high-resolution stratigraphy for the past 250 kyrs. The bottom-current paleospeed was inferred from a correlation of particle size in seafloor samples with mean current speed from nearby current-meters. The mean bottom-current speed at depths comparable to modern AABW was highest (7–10 cm/s) during interglacial to glacial transitions corresponding to the oxygen isotopic stage 6/7 and 4/5 boundaries and at present. The mean bottom-current speed at depths comparable to modern NADW was nearly uniform for most of the past 250 kyrs except during glacial oxygen isotopic stage 2 when the speed dropped to 2 cm/s, or one-half of the present speed. The application of the “calibrated” particle-size method to examine bottom-current paleospeed allows testing of paleoceanographic models which rely on assumptions or inferences of changes in bottom-water production rate during the late Pleistocene paleoclimatic fluctuations.

Neogene stratigraphy of the submerged U.S. Atlantic margin

Thirty boreholes and several hundred sea-floor samples provide a detailed but incomplete record of Neogene strata and depositional environments along the submerged part of the United States Atlantic margin. Three major sedimentary basins, the Blake Plateau Basin, the Baltimore Canyon Trough, and the Georges Bank Basin, contain Neogene sedimentary prisms as thick as 1200 m, comprising bathyal marine to nonmarine strata. Middle Miocene rocks compose the most widely represented and thickest unit (> 600 m in the Baltimore Canyon Trough). Calcareous microfossils provide excellent biostratigraphic and paleoenvironmental indicators except in some sparsely fossilferous shallow-water intervals and where shallow-water diatomaceous assemblages predominate. However, in outer sublittoral assemblages of the Blake Plateau Basin and Baltimore Canyon Trough, the co-occurrence of abundant diatoms, radiolarians, and planktic foraminifera provide an excellent opportunity for intercalibration of biozones. Preliminary correlations between chronostratigraphic and paleoenvironmental interpretations indicate a close association of depositional sequences with sea-level fluctuations, but some local exception persist.

Geophysical and Geologic Studies of Ross Sea Continental Margin, Antarctica: ABSTRACT

In February 1984, the U. S. Geological Survey conducted geophysical and geologic investigations of the Ross Sea Shelf and outer continental margin of Antarctica aboard the research vessel S. P. Lee. The geophysical data included 24-channel seismic-reflection, high-resolution seismic-reflection, sonobuoy seismic, gravity, magnetic-gradiometer, bathymetry, and heat-flow measurements. Sea-floor samples were collected for geologic and geochemical studies, using 3-m gravity corer, box corer, and rock dredge. Principal survey areas were along the southern and western Ross Shelf, near Cape Adare and Islin Bank, and along the central outer continental margin. These areas lie adjacent to those covered by earlier multichannel-seismic-reflection surveys made by France, West Germany, and Japan. Three north-south-trending sedimentary basins, containing as much as 5 km (16,404 ft) of Cenozoic sediment, lie beneath the Ross Sea Shelf and extend seaward beyond the continental shelf edge. These basins are separated by basement ridges and are bounded on the west by the Transantarctic mountains and on the east by mountain ranges of Marie Byrd Land. Seismic stratigraphy, crustal-thickness measurements, and rock samples from the Ross Sea region indicate that the three basins may have formed initially by crustal rifting during the middle and Late Cretaceous time and have subsequently filled with early Tertiary(?) as well as Oligocene and younger glacial marine sediment. Sedimentary thicknesses, heat-flow values, and geochemical analyses indicate that some parts of the Ross Sea Shelf m y have favorable conditions for the generation of hydrocarbons. End_of_Article - Last_Page 465------------

Deep Western Boundary Undercurrent delineated by sediment texture at base of North American continental rise

A Deep Western Boundary Undercurrent (DWBUC) flows south along the North American continental slope and rise between 1,000 m and 5,000 m with the exact location differing between investigators. The silt-mean particle-size of seafloor samples, however, defines a ∼300-km wide highvelocity zone within the DWBUC below ∼4,500 m. Analysis of silt subfractions shows that the DWBUC high-velocity zone winnows both fine and very fine silt, with preferential deposition of medium and coarse silt. Silt textural analyses indicate that velocity increases with depth in the DWBUC, reaching a maximum at 4,900 to 5,000 m consistent with results of the High Energy Benthic Boundary Layer Experiment (HEBBLE).

Sedimentology of Southwestern Roads Region, U.S. Virgin Islands-Origin and Rate of Sediment Accumulation

Sand deposits on southern insular shelf of St. Thomas, U.S. Virgin Islands, were investigated to determine their origin, environmental processes and accumulation rates. Sea-floor samples show that the sand has been derived (in situ) mainly from calcareous algae and molluscs. Zonation of the dominant sand producers is related to the present environmental setting; water depth has the greatest influence. Carbon-14 data (bulk sample) of cores indicate accumulation rates of slightly less than 1 mm/year for the last 5,000 years. Faunal studies show that the climate has remained constant over the past 5,000 years. The only changes in environmental conditions appear to have been an increase in water depth, changes in the patterns of water movement, and an increase in water temperature.--Modified journal abstract.

Oxygen and carbon isotope evidence for seawater-hydrothermal alteration of the Macquarie Island ophiolite

Rocks of the Miocene Macquarie Island ophiolite, south of New Zealand, have oxygen and carbon isotopic compositions comparable to those of seafloor rocks. Basalt glass and weathered basalts have δ 18O values at 5.8–6.0‰ and 7.9–9.5‰, respectively, similar to drilled seafloor rocks including samples from the Leg 29 DSDP holes near Macquarie Island. Compared to the basalt glass, the greenschist to amphibolite facies metaintrusives are depleted in 18O (δ 18O=3.2–5.9‰) similar to dredged seafloor samples, whereas the metabasalts are enriched (δ 18O=7.1–9.7‰). Although the gabbros are only slightly altered in thin-section they have exchanged oxygen with a hydrothermal fluid to a depth of at least 4.5 km. There is an approximate balance between 18O depletion and enrichment in the exposed ophiolite section. The carbon isotopic composition of calcite in the weathered basalts (δ 13C=1.0–2.0‰) is similar to those of drilled basalts, but the metamorphosed rocks have low δ 13C values (−14.6 to 0.9‰). These data are compatible with two seawater circulation regimes. In the upper regime, basalts were weathered by cold seawater in a circulation system with high water/rock ratios (≫1.0). Based on calcite compositions weathering temperatures were less than 20°C and the carbon was derived from a predominantly inorganic marine source. As previously suggested for the Samail ophiolite, it is postulated that the lower hydrothermal regime consisted of two coupled parts. At the deeper levels, seawater circulating at low water/rock ratios (0.2–0.3) and high temperatures (300–600°C) gave rise to 18O-depleted gabbro and sheeted dikes via open system exchange reactions. During reaction the seawater underwent a shift in oxygen isotopic composition (δ 18O=1.0–5.0‰) and subsequently caused 18O enrichment of the overlying metabasalts. In the shallower part of the hydrothermal regime the metabasalts were altered at relatively high water/rock ratios (1.0–10.0) and temperatures in the range 200–300°C. The relatively low water/rock ratios in the hydrothermal regime are supported by the low δ 13C values of calcite, interpreted as evidence of juvenile carbon in contrast to the inorganic marine carbon found in the weathered basalts.