Samail Ophiolite plutonic suite: Field relations, phase variation, cryptic variation and layering, and a model of a spreading ridge magma chamber

Geologic mapping of an intact plutonic sequence within the Samail ophiolite in the Ibra area, southeastern Oman Mountains, reveals stratigraphic, structural, and petrologic details of oceanic layer 3. Four measured stratigraphic sections, each spaced about 5 km apart across the southern flank of Jabal Dimh, define a time‐transgressive progression within the ophiolite and reveal geometric and petrologic features of a spreading‐ridge magma chamber. The sections show the following vertical sequence: (1) dunite (chr‐ol cumulates ± harzburgite xenoliths) 0–200 m thick, grading up from a transition zone with harzburgite tectonite, (2) interlayered wehrlite‐melagabbro‐gabbro (cpx‐ol and ol‐cpx‐pl cumulates) 0–100 m thick, (3) layered gabbro (chiefly ol‐cpx‐pl cumulates but including recurrent intervals of cumulus wehrlite and melagabbro) 2.6–5.5 km thick, (4) planar laminated nonlayered gabbro (chiefly ol‐cpx‐pl cumulates) 100–400 m thick, (5) hypidiomorphic (ol)‐hb‐cpx gabbro (high‐level gabbro) 200–800 m thick, (6) small, discontinous diorite to plagiogranite bodies at or near the top of the gabbro. Cumulus textures (adcumulus > mesocumulus), planar lamination, and cumulus layering (phase, ratio, and grain size layers at mm to 10‐m scale, commonly graded) within this sequence show that crystals accumulated from the base of the magma chamber upward to within a few hundred meters of the top; downward solidification from the roof was minor. Cyclicity within the cumulus sequence is represented by the recurrence of olivine‐rich intervals (melagabbo and wehrlite) up to high stratigraphic levels and by hundreds of phase‐graded layers (ol‐rich at the bases to pl‐rich at the tops), individually up to 5 m thick. Limited cryptic variation relative to closed‐system layered intrusions and the limited range in solid‐solution components of olivine (Fo 69–90 ), plagioclase (An 62–95 ) and clinopyroxene (En 40–54 , Fs 4–16 , Wo 37–49 ) from the cumulus suite require replenishment of the magma during its crystallization history. Zig‐zag normal and reverse cryptic variation differentiation trends (both in major solid‐solution components and in minor element concentration) indicate that the replenishment took place in pulses followed by periods of magma mixing and crystal fractionation. Recurrent olivine‐rich intervals commonly (but not always) coincide with reverse cryptic variation trends. They are considered to be the products of fresh draughts of primitive magma, with olivine and chromite as the only liquidus phases prior to extensive mixing with the more fractionated resident magma within the chamber. The sequence of crystallization is explained using the simplified tholeiitic basalt tetrahedron of Presnall et al. (1979). The plutonic sequence is roofed by sheeted dikes that are overlain by submarine basalt, indicating that the magma crystallized beneath a spreading ocean ridge. Major and trace element geochemistry of the dike complex is similar in many respects to that of mid‐ocean ridge basalt (MORB) and yield Mg/Mg + Fe +2 ratios that overlap with ratios predicted for parent liquids of the cumulus suite based on crystal/liquid equilibria. Field evidence, including continuity of layering over large areas, distinctive layer sets that are mapped for several km along strike, and lack of chamber edge contacts indicates that crystallization of Jabal Dimh plutonics took place in a single, large, long‐lived magma chamber rather than in small transitory chambers. Similarity of mineral compositions and tentative correlation of cryptic variation trends between stratigraphic sections favors a large, single‐chamber model. The chamber shape, controlled by floor growth upward greatly exceeding roof growth downward as the chamber halves diverged from the paleo‐spreading axis, must have been funnel‐shaped in cross section; i.e., the floor sloped inward to the center from a sandwich horizon that developed just beneath the roof at both sides. This conclusion is based on the occurrence of cumulates up to a very high level in the stratigraphic sections and on geologic mapping which shows that cumulus layering locally approaches the base of the cumulus section at an angle. This vergence angle is <20°, well below inferred angles of repose for cumulates. Tentative correlation of cryptic variation between sections yields similar vergence angles. The scarcity of slump structures and of layer disruption within the main layered sequence verifies that the floor sloped rather gently, surely much less than 45°. Thus, the half‐width of the chamber must have exceeded its vertical dimension, which averages ∼5 km. A half‐width of ∼18 km is calculated from a preferred vergence angle of 15°. The cumulates crystallized within this spreading chamber from a magma body that was repeatedly replenished by primitive magma from a central feeder zone at the spreading axis. High‐level gabbro and diorite through plagiogranite crystallized from stagnant residual magma at the outer edges of the chamber, beyond reach of replenishment from the center.