Fluid composition in subduction zones.
* the porosity of the originally water-rich sediments of accretionary complexes is rapidly reduced by tectonic forces,
* heat flow is regionally variable,
* depth profiles have characteristic temperature and pore-fluid chemical and isotopic anomalies that can only be maintained by rapid and rather recent fluid flow, and
* diffusive and/or channelized fluid venting is widespread.
The latter occurs along sedimentary and structural-tectonic conduits such as unconformities, faults, and the decollement (the prominent boundary between the overriding and underthrusting plates shown in the figure on page 87) as well as through mud volcanoes.
These fluids sustain prolific benthic biological communities and cause widespread carbonate deposition as cement, vein filling, crusts, or chimneys, mostly from oxidation of microbially or thermogenically derived methane. The fluids also play an important role in the deformational, thermal, and chemical evolution of subduction zones, and enhance sediment diagenesis and rock metamorphism. Fluids released from these reactions transport dissolved components into the ocean, some of which may be important for global geochemical budgets. At greater depths (more than 80 kilometers), released fluids, especially water from the subducted sediments and altered oceanic basement and carbon dioxide from methane oxidation and decarbonation reactions, may expedite partial melting processes in the overlying mantle wedge, leading to arc volcanism.
The presence of sediment-derived isotopes and trace elements, especially cosmogenic beryllium 10 (half life 1.5 million years) in arc lavas, provides evidence for sediment recycling in some subduction zones. Global estimates of sediment contribution to arc lavas range from a few to 20 percent of the subducted sediments.
The total volume of the internally available fluid sources in subduction zones through steady-state processes has been estimated to be 1 to 2 cubic kilometers per year. These estimates, however, do not account for the 2 to 6 order of magnitude larger than predicted fluid-flow rates measured at numerous channelized fluid venting sites, for example, at the Barbados, Nankai, and Cascadia accretionary complexes. This discrepancy in fluid volumes suggests either that the channelized fluid flow is transient in nature and/or that a major external fluid source exists. Meteoric water (rain or snow) is the most likely external source, but how it might be transported to the subduction zones is yet unknown.
Geochemistry of the Fluids
Detailed studies of the chemical and isotopic compositions, mostly of the pore fluids obtained through drilling and of the channelized venting fluids obtained with submersibles and conventional coring, indicate that the chemical and isotopic characteristics of the expelled fluids differ markedly from seawater, the original pore fluid. Of particular interest are the ubiquitous fresher-than-seawater fluids often found in accretionary complexes and associated with fluid conduits such as faults, the decollement, or mud volcanoes. Seawater chloride dilution of 10 to 64 percent has been recorded. Unraveling the origin of fresher-than-seawater fluids is of great importance to understanding subduction zone hydrogeochemistry. The only internal sources and processes that may provide water for the formation of the low-chloride fluids are: 1) Dehydration or breakdown of hydrous minerals, particularly clay minerals, amorphous opal (opal-A), and zeolites in the accretionary complex and of minerals such as talc, phengite, serpentine, and amphiboles in the oceanic basement, 2) Dissociation of gas hydrates (clathrates), ice-like crystalline compounds whose expanded ice-lattice forms cages that contain gas molecules (mostly methane hydrate has been recovered from several accretionary complexes, and geochemical and geophysical evidence for the presence of gas hydrates has been observed at most of them), and 3) Clay membrane ion filtration: Geochemical evidence for the occurrence or importance of the latter process in clay-rich subduction zones is yet unavailable.
These overall dilute and fresher-than seawater fluids are often characterized by other chemical and isotopic anomalies. They are generally enriched in alkalinity, lithium, sodium, silica, beryllium, boron, iodine, methane (ethane, propane), carbon dioxide, and hydrogen sulfide; in contrast, they are depleted in potassium, magnesium, and sulfate. Concentrations of calcium and strontium vary, influenced by carbonate recrystallization and, at greater depths, by decarbonation. Strontium isotopic ratios that vary from highly radiogenic continental crustal to nonradiogenic oceanic basement values suggest communication with various deep-seated basement sources. This is also supported by the presence of mantle-derived helium; for example, based on helium isotopic analyses, at Nankai below the decollement, about 25 percent of the helium is mantle-derived. Helium, like chloride, is an excellent geochemical tracer because it is conservative and unaffected by chemical or biological reactions; its isotopic composition uniquely defines its source. Trace amounts of magmatic methane and carbon dioxide may be present as well. If so, they would be masked by the abundant microbially and thermogenically-derived biogenic methane and carbon dioxide.
Unusually high pH (alkaline) and chloride-depleted (57 percent seawater dilution) fluids that are rich in methane, ethane, and propane as well as in hydrogen sulfide, carbonate alkalinity, and ammonia, have been recovered from the Conical Seamount, an active low-density serpentinite mud volcano in the Mariana forearc, and in the Chile convergent margin adjacent to the triple junction. This suggests a rather deep (greater than or equal to 10-kilometers) source for these fluids. The global flux of these unusual fluids is as yet unknown. The Mariana subduction zone lacks an accretionary complex; here all the sediment is being subducted.
A variety of seafloor bathymetric features known as mud volcanoes typify sites of focused fluid venting. Their shapes range from conical with diameters of a few meters to about 1 kilometer, to linear ridges that are occasionally greater than 10 kilometers long. These features are common at subduction zones; they have been found at every convergent plate margin surveyed at the appropriate resolution. One of the most extensively studied mud-volcano fields occurs at the Barbados convergent margin. Here the surface area of the mud volcanoes covers nearly 45 percent of an approximately 1,600-square-kilometer region. About 30 square kilometers of this mud-volcano field, situated on the oceanic side of the deformation front, has been extensively surveyed using geophysical and coring techniques.
Deep-towed side-scan sonar images of the Barbados margin show a variety of mud volcano features. These images allow us to record temperature gradients in detail and recover cores from individual mud volcanoes. One of the larger mud volcanoes is shown in the figure at left. In its center, temperatures of about 20 |degrees~ C were measured at only 1 meter below the sea floor (mbsf); the surrounding bottom water temperature is about 2 |degrees~ C. The venting fluid is characterized by chloride concentrations of 211 millimoles, about 40 percent of seawater's value. These temperatures and chloride concentrations reflect extraordinarily rapid, focused, vertical flow of fluid from the mud volcano. Numerical calculations based on the temperature gradients indicate flow rates of 17 meters per year. The temperature gradients and chloride dilution decrease closer to the edge of the volcano, indicating that flow is most rapid in its center.
Saline Fluids in Subduction Zones
The residual fluids from gas hydrate formation and clay membrane ion filtration are brines, fluids of high solute concentrations and of high density. However, brines have not been observed in association with the ubiquitous gas hydrates in accretionary complexes. This is best explained by loss of solutes through diffusion or fluid advection at the sites of hydrate formation.
In addition to precipitation or dissolution of evaporite minerals such as halite or sylvite, brines with more than twice seawater chloride concentrations result from the hydration of volcanogenic sediments or of oceanic basement rocks to hydrous minerals such as clay minerals and zeolites. An excellent example of a brine formation from seawater evaporation has been observed in the Peru forearc basins; brines from volcanic ash alteration occur in the New Hebrides intra-arc Aoba basin and in an Izu-Bonin forearc basin. At all the previously drilled accretionary complexes, however, the fresher-than-seawater fluids dominate. Saline fluids should, however, be present in accretionary complexes associated with evaporites, for example in the Mediterranean Sea. The only fluids with somewhat (less than 15 percent) higher chloride concentrations than seawater were observed in association with volcanogenic sediments in the Nankai and New Hebrides accretionary complexes. Similar to submarine hydrothermal fluids, the elevated chloride fluids associated with volcanogenic sediments or oceanic basement could evolve into calcium-chloride brines. Saline fluid inclusions have been observed in mineralized veins in metamorphic rocks of accretionary complexes. Because of the scarcity of geochemical data on these fluid inclusions, it is premature to speculate on their origin or quantitative importance.
Miriam Kastner, the first woman professor at Scripps Institution of Oceanography, gradually migrated westward from Harvard University where she received her Ph.D., through the University of Chicago at which she spent a year as a post doctoral fellow. During her first summer as a Harvard graduate student, she became interested in oceanography, and worked with a prominent conservative scientist on the geochemistry and mineralogy of sediments recovered from the flanks of the Mid-Atlantic Ridge. She is interested in natural-fluid rock processes, especially between seawater and marine sediments and oceanic basement. Her finite "spare" time is mostly dedicated to music.
Jonathan B. Martin came to Scripps Institution of Oceanography to work on fluids after graduating with a master's degree in geology from Duke University, where he worked on rocks. During the past year, he has completed his Ph.D. dissertation and produced a son, Peter. He is grateful to his wife, Ellen, for her help in both endeavors. After graduation, he will continue to work on convergent margin processes, both modern and ancient, at the University of California, Santa Cruz, and at the US Geological Survey.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||25 Years of Ocean rilling|
|Author:||Kastner, Miriam; Martin, Jonathan B.|
|Date:||Dec 22, 1993|
|Previous Article:||Studying crustal fluid flow with ODP borehole observatories.|
|Next Article:||Scientific ocean drilling and continental margins: understanding the fundamental transition from continent to ocean.|