Hydrothermal fluid flow through gabbros at IODP site 1309, Atlantis Massif, Mid-Atlantic Ridge

Alteration of the oceanic crust by hydrothermal circulation is one of the most important processes in the Earth System, responsible for cooling of ocean lithosphere affecting the heat budget of ocean crust and making a significant contribution to the chemistry and the isotopic composition of oceans. Oceanic crust has been proven to be structurally heterogeneous depending on the rate of spreading. Fast spreading crust is characterised by a layer cake model with pillow lavas underlain by sheeted dykes and plutonics, whereas ultra-slow/slow spreading oceanic crust is more complex with gabbro bodies intruded into peridotites brought to the surface via low angle detachment faults, forming oceanic core complexes (OCCs). Large hydrothermal systems such as TAG (Trans-Atlantic Geotraverse) are associated with detachment faults and may involve much deeper fluid circulation than typical systems at fast spreading ridges. The Atlantis Massif, 300N is an OCC located at the inside corner high between the Mid-Atlantic Ridge and the Atlantis Transform Fault that has been drilled during lOOP legs 304 and 305. lOOP Hole U1309B and 0 are dominated by gabbro with minor interlayered ultramafic rocks and diabases intruded at the top. The aims of this project are to characterise the fluids that circulated in the section sampled by lOOP Holes U1309B and 0 and to assess what processes control the fluid chemistry, to assess fluid fluxes related to various stages of alteration, and to better constrain how fluids circulate in OCCs by placing the results in the context of models for hydrothermal circulation at TAG. Fluid inclusion analyses - microthermometry and LAICPMS - and isotopic analyses of strontium and oxygen were undertaken in an attempt to answer these questions. Fluid inclusion microthermometry underlines the occurrence of four types of fluid in the Atlantis Massif. Fluid type 1 a is a seawater-like salinity fluid that is observed in late quartz vein precipitating at low pressure low temperature. Fluid type 1 b is depleted with respect to seawater salinity and is observed in plagioclase of gabbros and is the result of mixing with recharge seawater and supercritically phase-separated seawater-derived fluid. High salinity fluid (Type 3a) and halite-saturated fluid (type 3b) are observed in quartz grains of a trondjhemite intrusion. These fluids are interpreted to be generated by condensation of a magmaticv fluid. Fluid chemistry is controlled by phase separation processes and mainly by fluid-rock interactions. Isotopic analyses show that fluids circulated mainly close to the detachment fault and that limited amounts of fluid escaped into the footwall. Whole rock isotopic analyses show that gabbros are relatively little altered while serpentinites show elevated strontium isotope ratios. Small sample analyses show that gabbros are heterogeneous, with amphibole vugs and prehnite showing elevated seawater-like values, amphiboles replacing pyroxene intermediate values, and plagioclase commonly retaining igneous values. Serpentinites might be contaminated by late carbonate precipitation. However, the elevated strontium isotope ratio of prehnite replacing plagioclase during formation of micro-rodingite argues for the serpentinising fluid being seawater like. Oxygen isotope analyses support the conclusions of metamorphic petrology, that the majority of alteration took place at temperatures > 300 QC. The patterns of hydrothermal alteration can be understood in terms of kinetically limited exchange of isotopes between fluid and rock. High flux pathways such as the amphibole vugs were formed at low effective Darnkohler numbers (ND), such that the amphibole reflects the fluid composition while the altered plagioclase in the vug walls have rock- dominated isotopic ratios. Tremolite-talc veins also appear to have formed under high flux, low ND conditions, while tremolite-chlorite coronas and micro-rodingite veins are also quite high flux features. Reaction permeability may have played a role in generating all of these fluid pathways. Although nominal fluid fluxes can be calculated on the basis of the down hole isotopic profile, it is likely that the main direction of fluid flow was parallel to the fault and hence perpendicular to the Hole. The evolution of fluid flow and alteration in the Atlantis oee can be interpreted in terms of the TAG model in which fluid discharge at black smoker temperatures occurs up the fault zone.