The Lost City hydrothermal system: Constraints imposed by vent fluid chemistry and reaction path models on subseafloor heat and mass transfer processes

Since the first reported discovery of the Lost City hydrothermal system in 2001, it was recognized that seawater alteration of ultramafic rocks plays a key role in the composition of the coexisting vent fluids. The unusually high pH and high concentrations of H₂ and CH₄ provide compelling evidence for this. Here we report the chemistry of hydrothermal fluids sampled from two vent structures (Beehive: ~90-116°C, and M6: ~75°C) at Lost City in 2008 during cruise KNOX18RR using ROV Jason 2 and R/V Revelle assets. The vent fluid chemistry at both sites reveals considerable overlap in concentrations of dissolved gases (H₂, CH₄), trace elements (Cs, Rb, Li, B and Sr), and major elements (SO₄, Ca, K, Na, Cl), including a surprising decrease in dissolved Cl, suggesting a common source fluid is feeding both sites. The absence of Mg and relatively high concentrations of Ca and sulfate suggest solubility control by serpentine-diopside-anhydrite, while trace alkali concentrations, especially Rb and Cs, are high, assuming a depleted mantle protolith. In both cases, but especially for Beehive vent fluid, the silica concentrations are well in excess of those expected for peridotite alteration and the coexistence of serpentine-brucite at all reasonable temperatures. However, both the measured pH and silica values are in better agreement with serpentine-diopside-tremolite-equilibria. Geochemical modeling demonstrates that reaction of plagioclase with serpentinized peridotite can shift the chemical system away from brucite and into the tremolite stability field. This is consistent with the complex intermingling of peridotite and gabbroic bodies commonly observed within the Atlantis Massif. We speculate the existence of such plagioclase bearing peridotite may also account for the highly enriched trace alkali (Cs, Rb) concentrations in the Lost City vent fluids. Additionally, reactive transport modeling taking explicit account of temperature dependent rates of mineral dissolution and precipitation clarifies the feedback between permeability, heat loss, and changes in the dissolved Si of the vent fluids. Assuming both the Beehive and M6 vent fluids were sourced at similar subseafloor conditions (tremolite buffered at 200°C), model results indicate loss of approximately 30% Si upon cooling to ~150°C during upflow. However, Si concentrations remained largely conservative with continued cooling to lower temperatures owing to unfavorable reaction kinetics. While consistent with the Beehive endmember composition, these results fail to explain the relative Si depletion in the lower temperature M6 fluids. Thus, it may be that more robust kinetic models for silicates are needed to accurately account for the mechanism and rate of silica removal in the unusually high pH of the Lost City vent fluids.