The Jormua Ophiolite exposes a unique fragment of Red Sea -type oceanic crust formed in a continental break-up related setting at 1.95 Ga, and subsequently evolved at the outer edge of a non-volcanic continental margin until its final obduction ca. 50 Ma later. The basaltic lid of the Jormua Ophiolite is rather thin: mafic ultramafic cumulates are absent, and sheeted dyke complexes and extrusive units are less than 1000 m and 400 m thick, respectively. Locally, lavas seem even to have erupted directly onto mantle peridotites exposed on the sea floor.
Two distinct types of basalts have been recognised. The "early dykes", which are intrusive to the mantle tectonite, have fractionated (H)REE patterns, OIB-like mantle normalised spidergrams, low Zr/Nb (×6), and eNd(1.95 Ga)× -0.6, indicative of derivation from a deep enriched source. All lavas and remaining dykes belong to the second, EMORB -like "main suite". Most "main suite" samples cannot be related by fractional crystallization from a common parental magma. Most likely, individual samples represents distinct melt fractions that underwent only minor pre-eruptive crystal fractionation. Thus, the chemical composition of the basalts provides indirect evidence that large plutonic mafic-ultramafic crustal units never developed in Jormua, and that their absence is not a matter of selective preservation. Sheeted dykes are largely free of postmagmatic compositional changes, but most lava samples have been leached of Fe and Mg by seafloor alteration with concomitant enrichment in insoluble Si and Al. The "main suite" has a relative large range in eNd(1.95 Ga) from -1.1 to +2.9, and yields a poorly defined Sm-Nd isochron with a slope corresponding to an age of 1.72±0.12 Ga (e=+1.2, MSWD=10). This apparent age is significantly less than the U-Pb zircon age (1.95 Ga) of the Jormua Ophiolite. We believe, that this discrepancy is due to leaching of some pillow lava samples in LREE by CO2-rich metamorphic fluids, their former presence being proven by ubiquitous talc-carbonate alteration of the adjacent mantle peridotites. Those basaltic dykes that
are enclosed in serpentinised mantle peridotites remained unaffected by such fluids, but became "rodingised" during serpentinization, and later flushed by H2O-fluids derived from dehydrating serpentinites during regional heating.
The unaltered "main suite" samples are characterised by
flat REE patterns, variable Zr/Nb=6-17, chondritic Th/Ta, and only moderately depleted isotopic signatures [eNd(1.95 Ga)×+2.0]. Their composition can be explained by mixing a depleted source with a relatively uniform proportion of enriched component similar to that represented by the OIB-like "early dykes".
Later stages of continental rifting were probably characterised by magmatism from a deep source, witnessed not only by the OIB-like "early dykes", but also by coeval peralkaline granites that intruded the continent margin and now occur SSW from Jormua. Subsequently, as soon as the old subcontinental lithosphere was ruptured, the oceanic crust-forming "main suite" was generated from the upper asthenospheric source, which had been metasomatised by
the "early", OIB-like melts, or which gained such
enriched component by thermal erosion of the veined old subcontinental lithosphere.