C. Geoffrey Wheat

Hydrothermal circulation through mid-ocean ridge flanks: Fluxes of heat and magnesium☆

Thermally driven convection of seawater occurs through oceanic crust of all ages, at the seafloor spreading axis, on mid-ocean ridge flanks, and in the ocean basins. At the ridge axis and on the flanks, circulating seawater produces a major chemical exchange between the oceans and the crust. Based on heat budget constraints and the composition of hot springs, between 10 and 40% of the river flux of Mg can be taken up during high-temperature alteration of the basaltic crust along the ridge axis. Most of the hydrothermal heat loss, however, occurs on the mid-ocean ridge flanks, where the temperatures are lower and the seawater flux correspondingly larger. The estimated heat loss on the flanks is so large that upwelling must occur over a large fraction (5–30%) of the seafloor less than 65 Ma in age, if temperatures are < 20°C and seepage velocities are on the order of 10 to 100 cm/y. The circulating seawater needs to lose on average less than 1–2% of its Mg content in order to solve the Mg mass balance for the oceans. Chemical fluxes through mid-ocean ridge flanks are poorly known because of the wide range of crustal conditions that prevail there and the paucity of study to date. The most critical parameter for characterizing crustal conditions is temperature in basement, which is a function of crustal age, basement topography, and sediment thickness and permeability. Basement temperature, which can usually be inferred from heat flow and seismic reflection surveys, largely determines the change in the composition of seawater circulating through basement. This change can be inferred, in turn, from profiles of sediment porewater chemistry. All sites studied to date with temperatures at the sediment-basement interface ≤ 25°C have a large component of advective heat loss and show only a small ( 80%) Mg loss. Both types of sites may be important for chemical fluxes. Whether the cooler sites are important depends on how much the seawater changes in composition as it circulates through the crust; only small changes are needed. Whether the warmer sites are important depends on how much heat is lost by advection in this type of setting. The warmer sites could produce significant chemical fluxes even if they are scarce: they could account for the entire river input of Mg to the oceans even if they represent only 8–20% of the total advective heat loss on ridge flanks.

Composition of pore and spring waters from Baby Bare: global implications of geochemical fluxes from a ridge flank hydrothermal system

Warm hydrothermal springs were discovered on Baby Bare, which is an isolated basement outcrop on 3.5 Ma-old crust on the eastern flank of the Juan de Fuca Ridge. We have sampled these spring waters from a manned submersible, along with associated sediment pore waters from 48 gravity and piston cores. Systematic variations in the chemical composition of these waters indicate that hydrothermal reactions in basement at moderate temperatures (63°C in uppermost basement at this site) remove Na, K, Li, Rb, Mg, TCO2, alkalinity, and phosphate from the circulating seawater and leach Ca, Sr, Si, B, and Mn from the oceanic crust; and that reactions with the turbidite sediment surrounding Baby Bare remove Na, Li, Mg, Ca, Sr, and sulfate from the pore water while producing ammonium and Si and both producing and consuming phosphate, nitrate, alkalinity, Mn, and Fe. K, Rb, and B are relatively unreactive in the sediment column. These data confirm the earlier inference that sediment pore waters from areas of upwelling can be used to estimate the composition of altered seawater in the underlying basement, even for those elements that are reactive in the sediment column or are affected by sampling artifacts. The composition of altered seawater in basement at Baby Bare is similar to the inferred composition of 58°C formation water from crust nearly twice as old (5.9 Ma) on the southern flank of the Costa Rica Rift. The Baby Bare fluids also exhibit the same directions of net elemental transfer between basalt and seawater as solutions produced in laboratory experiments at a similar temperature, and complement compositional changes from seawater observed in seafloor basalts altered at cool to moderate temperatures. The common parameter among the two ridge flanks and experiments is temperature, suggesting that the residence time of seawater in basement at the two ridge-flank sites is sufficiently long for the solutions to equilibrate with altered basalt. This conclusion is supported by compilations of data from other ridge-flank sites, which show a systematic relationship between inferred basement water composition and temperature. We use the Baby Bare spring waters along with constraints from the riverine flux of Mg to estimate upper limits on the global fluxes of 14 elements at warm ridge-flank sites such as Baby Bare. Maximum calculated fluxes of Mg, Ca, sulfate, B, and K may equal or exceed 25% of the riverine flux, and such sites may represent the missing, high K/Rb sink required for the K budget. Additional fluxes from/to the crust on ridge flanks are expected from cool (<20–25°C) hydrothermal sites.

Phosphate removal by oceanic hydrothermal processes: An update of the phosphorus budget in the oceans

Abstract We present a compilation of dissolved phosphate and solid-phase P data from the oceanic crust to evaluate the effects of hydrothermal processes on the oceanic budget of P. Concentrations of phosphate in fluids that emanate from ridge-axis hydrothermal systems are less than that in bottom seawater. The extent of removal in these fluids is at least 30% and in some hydrothermal systems dissolved phosphate is removed completely from the circulating fluid. Evidence for the removal of phosphate in each of six ridge-flank hydrothermal systems is based on systematic variations in porewater profiles of phosphate and speeds of porewater flow. The extent of removal is >80% in these ridge-flank systems. These removal processes are recorded in the basaltic crust as an increase in P concentration that coincides with an increase in extent of alteration and content of ferric iron. Phosphate also is removed in hydrothermal plumes by coprecipitation with Fe oxyhydroxide particles, which eventually deposit on the seafloor. Each of these hydrothermal processes results in a flux of P into the oceanic crust. Bottom seawater flow through ridge-axis hydrothermal systems removes at most 0.4% of the preindustrial dissolved riverine flux of P, while ridge flanks remove at least 5%, but less than 50% of the dissolved riverine flux, consistent with P data from Deep Sea Drilling Project (DSDP) Sites 417 and 418. Removal of phosphate by coprecipitation with Fe-rich particles in hydrothermal plumes along ridge axes accounts for 18–33% of the dissolved riverine flux. Thus, hydrothermal systems remove about 50% of the preindustrial dissolved riverine flux of phosphate. We have included these new estimates in a revised budget for oceanic P.

Hydrothermal circulation, Juan de Fuca Ridge eastern flank: Factors controlling basement water composition

Pore water has been analyzed from sediment cores taken from three areas on the eastern flank of the Juan de Fuca Ridge as part of FlankFlux 90, a study of hydrothermal circulation through mid-ocean ridge flanks. Seismic reflection and heat flow surveys (Davis et al., 1992a) indicate that the three areas differ in sediment thickness, basement topography, abundance of outcrops, basement temperature, and fraction of heat lost by advection versus conduction. Area 1 is on 0.6 Ma crust with nearly continuous basement outcrop, area 2 is on 1.3 Ma crust over the first buried ridge parallel to the present ridge axis, and area 3 is on 3.5–3.8 Ma crust over two axis-parallel buried ridges that penetrate the sediment cover in three locations. Each area includes a hydrothermal system in which seawater flows into basement, reacts with crustal basalt, and then exits basement either through the sediment or directly into the overlying water column. As constrained by concentrations of sulfate and lithium in the pore waters, at least some seawater enters basement in all three areas without reacting fully with the overlying sediment, even where no outcrops are known nearby. Speeds of up welling of pore water through the sediment have been estimated by fitting profiles of dissolved magnesium and chlorinity, which behave conservatively in these areas, to numerical time-dependent transport models. The estimated velocities range from <0.1 to 7.4 cm/yr; faster flows probably occur but were not sampled. Upwelling speed correlates positively with heat flow and basement highs and negatively with sediment thickness. The correlation with heat flow differs from area 2 to area 3 along with differences in physical properties of the turbidite sediment. We have documented pore water upwelling through sediment up to 100 m thick. We estimate that upwelling continues at decreasing speeds through sediment up to 160 m thick, corresponding to a heat flow of 0.44 W/m2 in area 2 and 0.3 W/m2 in area 3. Concentrations of magnesium and chlorinity in the altered seawater upwelling from basement are uniform within each area but differ from one area to the next. Both species remain at the bottom seawater concentration in area 1, where basement is cooled to <10°C at the base of the sediments mainly by advection. The concentration of magnesium decreases with increasing basement temperature in areas 2 and 3 to a minimum of 2.5 mmol/kg at about 90°C in area 3. The transition from largely advective to largely conductive heat loss occurs over only 20 km between areas 1 and 2 and corresponds to a dramatic change in the composition of fluid circulating through basement, as the uppermost basement is heated from <10° to 40–50°C. Chlorinity of the basement fluid increases above the present-day bottom seawater concentration in areas 2 and 3 and in nearly all other mid-ocean ridge flanks studied to date, as a result of rock hydration and the higher chlorinity of bottom seawater during the last glacial period. While chlorinity generally correlates positively with uppermost basement temperature in various ridge flank hydrothermal systems, it reaches a maximum in area 2 at only 40°C, probably because alteration there occurs at a lower water/rock ratio than elsewhere. For all mid-ocean ridge flanks studied to date, the temperature at the basement interface correlates better with the fraction of heat lost by advection versus conduction and with the average thickness of the sediment cover than with crustal age.

Colonization of subsurface microbial observatories deployed in young ocean crust.

Oceanic crust comprises the largest hydrogeologic reservoir on Earth, containing fluids in thermodynamic disequilibrium with the basaltic crust. Little is known about microbial ecosystems that inhabit this vast realm and exploit chemically favorable conditions for metabolic activities. Crustal samples recovered from ocean drilling operations are often compromised for microbiological assays, hampering efforts to resolve the extent and functioning of a subsurface biosphere. We report results from the first in situ experimental observatory systems that have been used to study subseafloor life. Experiments deployed for 4 years in young (3.5 Ma) basaltic crust on the eastern flank of the Juan de Fuca Ridge record a dynamic, post-drilling response of crustal microbial ecosystems to changing physical and chemical conditions. Twisted stalks exhibiting a biogenic iron oxyhydroxide signature coated the surface of mineral substrates in the observatories; these are biosignatures indicating colonization by iron oxidizing bacteria during an initial phase of cool, oxic, iron-rich conditions following observatory installation. Following thermal and chemical recovery to warmer, reducing conditions, the in situ microbial structure in the observatory shifted, becoming representative of natural conditions in regional crustal fluids. Firmicutes, metabolic potential of which is unknown but may involve N or S cycling, dominated the post-rebound bacterial community. The archaeal community exhibited an extremely low diversity. Our experiment documented in situ conditions within a natural hydrological system that can pervade over millennia, exemplifying the power of observatory experiments for exploring the subsurface basaltic biosphere, the largest but most poorly understood biotope on Earth.

Trace element and REE composition of a low-temperature ridge-flank hydrothermal spring

Abstract Warm (25°C) hydrothermal springs have been sampled on Baby Bare, a basaltic outcrop on 3.5-Ma-old crust ∼100-km east of the Endeavor Segment of the Juan de Fuca Ridge. The source for these springs is a 62 to 64°C formation water that has cooled conductively as it ascends to feed the springs. This water originated as bottom seawater that probably descended into basement ∼52 km to the southwest at another, much larger outcrop called Grizzly Bare. As this seawater flows towards Baby Bare, it is heated and altered by reactions within basaltic basement and by diffusive fluxes to and from the overlying sediment. Concentrations of Mn, Co, Ni, Zn, Cd, and Mo in the spring waters are greater than in bottom seawater, indicating that the oceanic crust is a source for these elements to the oceans. At least a portion of this increase probably results from the redox cycling of Mn in sedimentary sources near the basement interface that produces a diffusive flux to basement formation waters. Additional removal of Mo and inputs of the other five elements to two of the three springs are observed locally near sites of venting, where density gradients can form shallow circulation cells within the sediment and diffusive exchange occurs. Concentrations of Cu, U, V, Y, and the rare earth elements (REEs, excluding Ce) in these samples are less than in bottom seawater, indicating that the oceanic crust is a net sink for these elements in this environment. Copper is probably removed into newly formed carbonate and/or sulfide phases. Removal of the oxyanions U and V is consistent with a net removal of phosphate demonstrated previously for ridge-flank hydrothermal systems. Similarly, removal of Y and the REEs is associated with carbonate, phosphate-rich, and oxide phases. Calculated maximum global chemical fluxes from “warm” ridge-flank hydrothermal systems such as Baby Bare are insignificant relative to riverine fluxes for these elements, except possibly for Mn and Mo. The impact on global geochemical budgets for these elements from lower temperature (

Chemical composition of basement fluids within an oceanic ridge flank : Implications for along-strike and across-strike hydrothermal circulation

Compositions of basement fluids are presented for four sites along a 3.5-m.y.- old, partly buried basement ridge on the eastern flank of the Juan de Fuca Ridge. This ridge is roughly parallel to the active ridge axis of the Endeavor Segment ∼100 km to the west. From south to north these sites are Baby Bare Outcrop, Ocean Drilling Program (ODP) Site 1026, and the southern and northern sides of Mama Bare Outcrop. The composition of basement fluids is determined or estimated from analyses of pore water samples that were extracted from sediments at each of these sites, spring waters from Baby Bare, and basement fluids that vented from the open ODP Hole 1026B. Chemical trends in basement fluids along this transect show increasing alteration from south to north. A similar trend was observed along an ODP transect perpendicular to the ridge axis with increasing fluid alteration from west to east. Much of the increase in fluid alteration along the ODP transect is explained by greater water-rock exchange with increasing basement temperature to the east. In contrast, the trend along the 3.5-m.y.-old ridge is best explained by diffusive exchange with the overlying sediment. The rate of this exchange is used to constrain hydrologie properties within basaltic basement. Flow within the 3.5-m.y.-old ridge is inferred to occur from south to north and lacks significant exchange with basement fluids from the active ridge crest to the west. Thus the two flow systems are hydrologically distinct, and flow paths are likely influenced by the complex distribution of permeability in basement, the pattern of seafloor morphology, and the type and rate of sedimentation.

Effect of fluid-sediment reaction on hydrothermal fluxes of major elements, eastern flank of the Juan de Fuca Ridge

On the eastern flank of the Juan de Fuca Ridge, reaction between upwelling basement fluid and sediment alters hydrothermal fluxes of Ca, SiO2(aq), SO4, PO4, NH4, and alkalinity. We used the Global Implicit Multicomponent Reactive Transport (GIMRT) code to model the processes occurring in the sediment column (diagenesis, sediment burial, fluid advection, and multicomponent diffusion) and to estimate net seafloor fluxes of solutes. Within the sediment section, the reactions controlling the concentrations of the solutes listed above are organic matter degradation via SO4 reduction, dissolution of amorphous silica, reductive dissolution of amorphous Fe(III)-(hydr)oxide, and precipitation of calcite, carbonate fluorapatite, and amorphous Fe(II)-sulfide. Rates of specific discharge estimated from pore-water Mg profiles are 2 to 3 mm/yr. At this site the basement hydrothermal system is a source of NH4, SiO2(aq), and Ca, and a sink of SO4, PO4, and alkalinity. Reaction within the sediment column increases the hydrothermal sources of NH4 and SiO2(aq), increases the hydrothermal sinks of SO4 and PO4, and decreases the hydrothermal source of Ca. Reaction within the sediment column has a spatially variable effect on the hydrothermal flux of alkalinity. Because the model we used was capable of simulating the observed pore-water chemistry by using mechanistic descriptions of the biogeochemical processes occurring in the sediment column, it could be used to examine the physical controls on hydrothermal fluxes of solutes in this setting. Two series of simulations in which we varied fluid flow rate (1 to 100 mm/yr) and sediment thickness (10 to 100 m) predict that given the reactions modeled in this study, the sediment section will contribute most significantly to fluxes of SO4 and NH4 at slow flow rates and intermediate sediment thickness and to fluxes of SiO2(aq) at slow flow rates and large sediment thickness. Reaction within the sediment section could approximately double the hydrothermal sink of PO4 over a range of flow rates and sediment thickness, and could slightly decrease (by ≤10%) the size of the hydrothermal source of Ca.

Continuous sampling of hydrothermal fluids from Loihi Seamount after the 1996 event

For at least 9 years prior to July 1996, hydrothermal fluids flowed from Peles Vents on Loihi Seamount, Hawaii. In July–August 1996 a tectonic-volcanic event occurred that destroyed Peles Vents, creating a pit crater (Peles Pit) and several sites with hydrothermal venting. In October 1996 we deployed two new continuous water samplers (OsmoSamplers) at two of these hydrothermal sites and collected fluids using traditional sampling techniques to monitor the evolution of crustal and hydrothermal conditions after the event. The samplers were recovered in September 1997, and additional discrete vent fluid samples were collected. The OsmoSampler located along the south rift at Naha Vents captured a change in composition from a low-chlorinity, high-K fluid (relative to bottom seawater) to a high-chlorinity, low-K fluid. These changes are consistent with the fluid cooling during ascent and being derived from several different sources, which include high- (>330°C) and low- ( 330°C) into which magmatic volatiles were added. During the deployment, thermal and fluid fluxes decreased. At Naha the transport of heat and chemicals was decoupled. The chemical and thermal evolution of hydrothermal fluids after the event on Loihi is consistent with previous models based on events that have occurred along mid-ocean ridges. The event at Loihi clearly had an effect on the local hydrography; however, the integrated effect of chemical fluxes to global budgets from similar events is uncertain. Chemical fluxes from similar events may have a global impact, if ratios of chemical (e.g., CO2, Fe/Mn, Mg, sulfate, and K) to thermal anomalies greatly exceed, or are in the opposite direction to, fluxes from mid-ocean ridge hydrothermal systems.

Seawater transport and reaction in upper oceanic basaltic basement: chemical data from continuous monitoring of sealed boreholes in a ridge flank environment

Osmotically pumped fluid samplers were deployed in four deep-sea boreholes that were drilled during Ocean Drilling Program (ODP) Leg 168 on the eastern flank of the Juan de Fuca Ridge. Samplers were recovered from ODP Sites 1024 and 1027 and aliquots were analyzed for a variety of dissolved ions. Results from both of the samplers show a drastic change in the major ion composition within the first 20–40 days after the borehole was sealed at the seafloor followed by a more gradual change in composition. This gradual change ceased after 820 days at Site 1024 but continued throughout the 3-year deployment at Site 1027. We modeled this change in composition to estimate the flux of formation fluid through the open borehole. The rapid early change requires a flow of ∼1500 kg of formation fluid per day. The more gradual later change requires flow rates of 38 kg/day at Site 1024 and 17.5 kg/day at Site 1027. The latter fluxes require a minimum average specific discharge of meters to hundreds of meters per year through the surrounding basaltic matrix. Trace element data show surprisingly little contamination given the presence of steel casing, Li-organic-rich grease at each joint, cement, and drilling muds. Observed changes in trace element concentrations relative to those of bottom seawater provide a measure for the global significance of cool (23°C; ODP Site 1024) ridge flank hydrothermal systems relative to warm (64°C; Baby Bare and ODP Site 1027) hydrothermal systems and illustrate the importance of these cooler systems to global geochemical budgets.