Origin of mineral chemical stratification in the mantle section of the Table Mountain massif (Bay of Islands Ophiolite, Newfoundland, Canada)

• Published in: Lithos Vol. 31; no. 3; pp. 81 - 102
• Main Authors: Suhr, Günter, Robinson, Paul T.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 1994; Oxford Elsevier; New York, NY
• Subjects: Crystalline rocks; Earth sciences; Earth, ocean, space; Exact sciences and technology; Igneous and metamorphic rocks petrology, volcanic processes, magmas; North America; Newfoundland; Eastern Canada; major elements; spinel; orthopyroxene; ophiolite; igneous rocks; olivine; clinopyroxene; partial melting; ultramafics; electron probe data; peridotites
• ISSN: 0024-4937; 1872-6143
• DOI: 10.1016/0024-4937(94)90002-7

Mineral chemical data have been collected from the 6 km thick mantle section of the Table Mountain massif (Bay of Islands Ophiolite Complex, Newfoundland). The samples chosen represent the geochemical background, i.e. exclude samples from the vicinity of bands, dykes and pods of dunites and pyroxenites within the host harzburgites and (rare) lherzolites. Three large mineral-chemical domains can be distinguished as a function of depth below the crustal gabbros. The intermediate mantle section (2–5 km beneath the gabbros) has a very depleted geochemical pattern with Cr# [ 100 Cr ( Cr + Al) ] in spinel > 44, low Ti. The basal kilometer of the mantle section shows a steep geochemical gradient towards less depleted mineral chemistries (Cr# as low as 10). Both the intermediate and basal mantle rocks fit into a progressive partial melting trend. The top of the mantle section (0–2 km below the gabbros) shows Cr# between 30 and 44, but also higher Ti and ferric iron in spinel, and Na in clinopyroxene. Microstructural evidence for melt-related clinopyroxene is locally preserved and, in one area, plagioclase lherzolites occur. The topmost mantle section is considered to have been affected by melt infiltration. The geochemical variation is larger than elsewhere, suggesting that melt-infiltration, or at least the process of trapping of melt, was heterogeneous. In most melt-infiltrated harzburgites, plagioclase is absent. Other peridotites show good microstructural evidence for melt-infiltration but little chemical evidence to substantiate it. Both features are discussed and can be explained by a variety of processes demonstrating the complexity of melt-infiltration.