The structure and geochemistry of the gabbro zone of the Annieopsquotch ophiolite, Newfoundland: implications for lower crustal accretion at spreading ridges

The Annieopsquotch ophiolite exposes a c. 5.5-km-thick section of tholeiitic gabbros, sheeted dykes and pillow basalts. Based on the along-strike consistency in thickness of the major crustal units, and lack of significant throw on spreading-related normal faults, the Annieopsquotch ophiolite is interpreted to have formed at an intermediate- to fast-spreading ridge. The upper c. 400 m of the gabbro zone is composed of massive, texturally heterogeneous gabbros with compositions that approach those of the sheeted dykes and basalts. Below this is c. 1.6 km of 10–30-m-thick planar intrusive sheets or sills. The lowermost part of the gabbro zone is composed of gabbroic rocks with relict troctolite and troctolitic gabbro enclaves, which are veined and partly replaced by gabbro and pyroxenite. Sill contacts within the central, sill-dominated part of the gabbro zone are sub-parallel to boundaries between the major ophiolite lithostratigraphic units. The upper and lower contacts of individual gabbro sills may have finer grain sizes, indicating that the intrusions cooled from the top and bottom. Locally well-preserved comb structures (crescumulates) indicate downward growth, supporting a sill interpretation. The sills are composed of weakly or un-deformed plagioclase+clinopyroxene ±olivine cumulates. Trace element modeling suggests that the parental magmas of these cumulates had compositions very similar to the overlying sheeted dykes and basaltic lavas, as do dykes emplaced within the gabbro zone. Model liquids calculated from the gabbroic sills generally become more evolved up-section, indicating that magma evolution in the Annieopsquotch ophiolite was dominated by fractionation in lower crustal conduits, below the level of a putative axial melt lens (AML). The model liquids, sheeted dykes and basalts preserve a similar, wide range of compositions, which may indicate that aggregation, homogenization and fractionation in an AML was inefficient. Similar intra-conduit fractionation of mantle-derived melts might also contribute to MORB evolution at mid-ocean ridges