The Samail nappe, Oman, is a classic ophiolite complex consisting of a thick section (~12 km) of "depleted" peridotite tectonite, overlain by layered gabbro cumulates (~5 km) and high-level noncumulate gabbro (< 1 km), followed by a 100% sheeted dike complex (~1.3 km), and capped by a series of highly altered pillow lavas (~0.5 km). The ophiolite stratigraphy is exposed in the Oman Mountains, an arid region covering over 15,000 km², with relief exceeding 2 km. In the first part of this thesis, the results of field mapping along a 20 x 125 km strip from Muscat to the Wahiba Sands are presented forming the foundation for the isotopic work reported in the second part of the thesis.
The upper mantle and oceanic crustal sections preserve a partial record of the events that occurred in the dynamic spreading environment underying the Hawasina Ocean, a portion of the Tethys Seaway. The Samail peridotite section records a history of partial melting, plastic flow, and reaction with transitory melts. Ascending melts reacted with the harzburgite wall-rocks leaving behind dunite (at all depths in conduits with high melt/rock), and olivine orthopyroxenite, websterite, clinopyroxenite, and gabbro (from deeper-to-shallower levels in conduits with low melt/rock). The entire peridotite section was affected by pervasive upper mantle deformation during its ascent, and a basal harzburgite-dunite zone may represent the boundary of the originally vertical conduit that fed both melts and peridotite to the Samail ridge system. The supply of melt was sufficient to produce an "open system" magma chamber that achieved a steady state with respect to cumulus phases; orthopyroxene saturation was rarely attained in the gabbro section. Crystal accumulation predominantly occurred from the bottom upward. Because the high-level gabbro intrudes > 90% of the overlying dike complex, both the diabase dikes and pillow lavas are interpreted to have been intruded close to the ridge-axis and to be comagmatic with the cumulate gabbro section. The thin, heterogeneous high-level gabbro preserves a record of many piecemeal stoping events occurring in a chamber roof environment that was mechanically unstable. Plagiogranite commonly occurs near the gabbro-diabase contact; field evidence and ¹⁸O/¹⁶O relationships demonstrate that the plagiogranites either form by partial melting of stoped blocks of hydrothermally altered roof rock, by extreme differentiation of a hydrous tholeiitic magma strongly modified by exchange and dehydration of such stoped blocks. The data suggest that the magma chamber was open both at the bottom and the top, and thus, MOR basalts are not considered to be primary, unmodified melts of the mantle.
The shallow-level magma chamber (< 3 km below the seafloor) was the heat engine that drove convective seawater circulation through joints and fractures in the overlying section of diabase and basalt and simultaneously in the gabbro underlying the "wings" of the funnel-shaped chamber. Both stable (¹⁸O/¹⁶O and D/H) and radiogenic (Sm/Nd and Rb/Sr) isotope systems were used to investigate the characteristics of the hydrothermal system. The Sm/Nd system was virtually unaffected by seawater-hyrothermal hydrothermal alteration and crystallization ages were obtained from plagioclase and pyroxene separates on single gabbro samples. Measured δ¹⁸O values in whole-rocks, 2.5 < δ¹⁸O < l9.6, from the oceanic crustal section are typically depleted in the lower parts of the section, and enriched in the upper parts, relative to the primary magmatic value of 5.7 ± 0.2. Also, the initial ⁸⁷Sr/⁸⁶Sr ratios vary from 0.7030 to 0.7065 increasing upward. The large variations in δ¹⁸O and ⁸⁷Sr/⁸⁶Sr are clearly the result of seawater hydrothermal alteration, demonstrating that large amounts of heated seawater (T > 500°C) penetrate deep into oceanic layer three as far down as the oceanic Moho. Mineral-mineral δ¹⁸O systematics (e.g. plagioclase-clinopyroxene) have been used to demonstrate isotope disequilibrium caused by this subsolidus hydrothermal exchange, to ascertain the primary δ¹⁸O values of the gabbro and plagiogranite reservoirs, to estimate relative exchange rates between minerals, to deduce the δ¹⁸O changes occurring in the hydrothermal fluids at various levels within the crust, and to differentiate between the effects of closed and open system hydrothermal exchange in natural systems.
The ¹⁸O redistribution within the oceanic crust was systematic, with whole-rock δ¹⁸O increasing from a minimum δ¹⁸O = 3.7 in the cumulate section about 1-2 km below the gabbro-diabase contact to values as high as 19.6 in the pillow lava section. A mass-balance calculation for the entire oceanic crustal section indicates that, the net change over the whole oceanic crustal section was zero implying that seawater also did not change and thus the seawater-oceanic crustal system was at some steady-state during the late Cretaceous. Modeling of the circulation extrapolated to the world-wide ridge system suggests that the δ¹⁸O of seawater is controlled by the hydrothermal interactions, and will be buffered to within 1 per mil of its present-day value as long as global spreading rates exceed 1 km²/yr.
