An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail Ophiolite, Oman: Evidence for δ^(18)O buffering of the oceans by deep (>5 km) seawater-hydrothermal circulation at mid-ocean ridges

摘要

Isotopic analyses of 75 samples from the Samail ophiolite indicate that pervasive subsolidus hydrothermal exchange with seawater occurred throughout the upper 75% of this 8-km-thick oceanic crustal section; locally, the H_2O even penetrated down into the tectonized peridotite. Pillow lavas (δ^(18)O = 10.7 to 12.7) and sheeted dikes (4.9 to 11.3) are typically enriched in ^(18)O, and the gabbros (3.7 to 5.9) are depleted in ^(18)O. In the latter rocks, water/rock ≤ 0.3, and δ^(18)O_(cpx) ≈ 2.9 + 0.44 δ^(18)O_(feld), indicating pronounced isotopic disequilibrium. The mineral δ^(18)O values approximately follow an exchange (mixing) trajectory which requires that plagioclase must exchange with H_2O about 3 to 5 times faster than clinopyroxene. The minimum δ^(18)O_(feld) value (3.6) occurs about 2.5 km below the diabase-gabbro contact. Although the gabbro plagioclase appears to be generally petrographically unaltered, its oxygen has been thoroughly exchanged; the absence of hydrous alteration minerals, except for minor talc and/or amphibole, suggests that this exchange occurred at T > 400°–500°C. Plagioclase δ^(18)O values increase up section from their minimum values, becoming coincident with primary magmatic values near the gabbro-sheeted diabase contact and reaching 11.8 in the diabase dikes. These ^(18)O enrichments in greenschist facies diabases are in part due to exchange with strongly ^(18)O-shifted fluids, in addition to retrograde exchange at much lower temperatures. The δ^(18)O data and the geometry of the mid-ocean ridge (MOR) magma chamber require that two decoupled hydrothermal systems must be present during much of the early spreading history of the oceanic crust (approximately the first 10^6 years); one system is centered over the ridge axis and probably involves several convective cells that circulate downward to the roof of the magma chamber, while the other system operates underneath the wings of the chamber, in the layered gabbros. Upward discharge of ^(18)O-shifted water into the altered dikes from the lower system, just beyond the distal edge of the magma chamber, combined with the effects of continued low-T hydrothermal activity, produces the ^(18)O enrichments in the dike complex. Integrating δ^(18)O as a function of depth for the entire ophiolite establishes (within geologic and analytical error) that the average δ^(18)O (5.7 ± 0.2) of the oceanic crust did not change as a result of all these hydrothermal interactions with seawater. Therefore the net change in δ^(18)O of seawater was also zero, indicating that seawater is buffered by MOR hydrothermal circulation. Under steady state conditions the overall bulk ^(18)O fractionation (Δ) between the oceans and primary mid-ocean ridge basalt magmas is calculated to be +6.1 ± 0.3, implying that seawater has had a constant δ^(18)O≈−0.4 (in the absence of transient effects such as continental glaciation). Utilizing these new data on the depth of interaction of seawater with the oceanic crust, numerical modeling of the hydrothermal exchange shows that as long as worldwide spreading rates are greater than 1 km^2/yr, ^(18)O buffering of seawater will occur. These conclusions can be extended as far back in time as the Archean (> 2.6 eons) with the proviso that Δ may have been slightly smaller (about 5?) because of the overall higher temperatures that could have prevailed then. Thus ocean water has probably had a constant δ^(18)O value of about −1.0 to +1.0 during almost all of earth's history.