Reconstructing the oxygen isotope composition of late Cambrian and Cretaceous hydrothermal vent fluid

• Published in: Geochimica et Cosmochimica Acta
• Main Authors: Turchyn, Alexandra V., Alt, Jeffrey C., Brown, Shaun T., DePaolo, Donald J., Coggon, Rosalind M., Chi, Guoxiang, Bédard, Jean H., Skulski, Thomas
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2013
• Subjects: 01 - Climate Change and Earth-Ocean Atmosphere Systems; Betts Cove; Canada; Newfoundland
• Online Access: http://eprints.esc.cam.ac.uk/3018/; http://eprints.esc.cam.ac.uk/3018/1/1-s2.0-S0016703713004596-main.pdf; http://www.sciencedirect.com/science/article/pii/S0016703713004596; https://doi.org/10.1016/j.gca.2013.08.015

Oxygen isotope analyses (δ18O) of 16 quartz–epidote pairs from late Cambrian (Betts Cove and Mings Bight, Newfoundland), Ordovician (Thetford Mines, Québec, Canada) and Cretaceous (Troodos, Cyprus) ophiolites are used to calculate the δ18O of the hydrothermal fluids from which they crystallized. We combine these with 3 quartz-fluid inclusion measurements and 3 quartz–magnetite measurements from the Cambrian ophiolites to explore how the range in the δ18O of submarine hydrothermal vent fluid has varied between the late Cambrian, Cretaceous and today. The range of calculated δ18O values of vent fluid (−4 to +7.4) is larger than that of modern seafloor hydrothermal vent fluid (0 to +4). We employ two numerical models to ascertain whether this range is most consistent with changes in paleo-seawater δ18O or with changes in the reactive flow path in ancient hydrothermal systems. A static calculation of the vent fluid oxygen isotope composition as a function of the water–rock ratio suggests that in an ocean with a lower δ18O than today, the range of vent fluid δ18O should be larger. Our data, however, show little evidence that the δ18O of the ocean was much lower than the global ice-free value of −1.2. A dual porosity model for reactive flow through fractured and porous media is used to model the relative evolution of the 87Sr/86Sr and δ18O of vent fluid in contact with rock. Our 87Sr/86Sr and δ18O for Cretaceous epidotes suggest the strontium concentration of the Cretaceous oceans may have been much higher than at present. The 87Sr/86Sr and δ18O data from Cambrian epidotes are strikingly different from the younger samples, and are difficult to model unless fluid-rock interaction in the Cambrian hydrothermal systems was substantially different. It is also possible that some of the quartz–epidote veins have been reset by obduction-related metamorphism. Our data suggest that the high calcium-to-sulfate ratio in early (and Cretaceous) seawater may have affected the degree of strontium isotope exchange, causing hydrothermal fluids to have 87Sr/86Sr closer to that of seawater than in modern systems.