Description: During Leg 163, in the Southeast Greenland margin, drilling penetrated a thick succession of volcanic rocks that erupted during breakup of the North Atlantic in the early Tertiary. Samples recovered during Leg 163 supplement the samples recovered during Leg 152 from the transect across the volcanic margin at 63°N. During Leg 163, drilling at Site 989 was intended to recover the oldest part of the continental prebreakup series, and drilling at Site 990 was intended to penetrate the transition zone from synbreakup, compositionally variable volcanic products, to postbreakup volcanics with a limited compositional range and a depleted chemical character similar to mid-ocean-ridge basalts (oceanic character). All the lava flows recovered from Sites 989 and 990, however, have an oceanic chemical character with low contents of incompatible elements and high contents of Sc. A dikelet from Site 990 and a previously drilled dike from Site 917 are likewise oceanic. We consider that the two drilled lava flows from Site 989 were emplaced after breakup despite their setting on the innermost part of the continental margin. The succession at Site 990 consists of 13 units of lava flow that show a slight compositional development upsection with Mg# decreasing from ~62 to ~49. This variation is within the limits known from the slightly younger oceanic succession drilled earlier at Site 918, in which the variation is thought to reflect fractionation in magma chambers in the oceanic crust. The reestablishment of magma chambers after the breakup must have been achieved during the interval that the short, undrilled lava succession between uppermost Site 917 and lowermost Site 990 was deposited. All the lava flows from Sites 989 and 990, except perhaps one (Unit 989-1), are crustally contaminated, as judged from their high Ba/Zr (〉0.42) ratios. The two dikes appear to be uncontaminated. The contamination most likely took place in magma chambers in the young oceanic rift, which, at this early stage of spreading, could still have contained fragments of continental crust. Most lavas were probably erupted within the rift and flowed subaerially away from it toward the edge of the continent. The dikes show that lateral injection of magma into the continental crust also took place without resulting in crustal contamination. The oceanic lava flows at Site 989 on the innermost margin have probably been erupted through such laterally injected dikes. The primary magma for the postbreakup rocks is estimated to have contained ~18% MgO. After fractionation of 30 mol% olivine (Fo91-82), it shifted to gabbro fractionation, and the magmas started to erupt. The modal composition of the gabbroic cumulate is 8% olivine (Fo82-74), 50% plagioclase (An76-66), and 42% clinopyroxene (Fs9-15). The erupted oceanic magmas are produced by 〈24 mol% gabbro fractionation, with an average of 14%. Thus, geochemical modeling indicates that the oceanic crust is composed of olivine cumulates (30%), gabbro cumulates (14%), and melts (lava flows and dikes, 56%), which is in accordance with models based on geophysical data.

Description: Voluminous, subaerial magmatism resulted in the formation of extensive seaward-dipping reflector sequences (SDRS) along the Paleogene Southeast Greenland rifted margin. Drilling during Leg 163 recovered basalts from the SDRS at 66ºN (Site 988) and 63ºN (Sites 989 and 990). The basalt from Site 988 is light rare-earth-element (REE) enriched (La(n)/Yb(n) = 3.4), with epsilon-Nd(t=60) = 5.3, 87Sr/86Sr = 0.7034, and 206Pb/204Pb = 17.98. It is similar to tholeiites recovered from the Irminger Basin during Leg 49 and to light-REE-enriched tholeiites from Iceland. Drilling at Site 989, the innermost of the sites on the 63ºN transect, was proposed to extend recovery of the earliest part of the SDRS initiated during Leg 152. These basalts are, however, younger than those from Site 917 and are compositionally similar to basalts from the more seaward Sites 990 and 915. Many of the basalts from Sites 989 and 990 show evidence of contamination by continental crust (e.g., epsilon-Nd(t=60) extends down to -3.7, 206Pb/204Pb extends down to 15.1). We suggest that the contaminant is a mixture of Archean granulite and amphibolite and that the most contaminated basalts have assimilated ~5% of crust. Uncontaminated basalts are isotopically similar to basalts from Site 918, on the main body of the SDRS, and are light-REE depleted. Consistent with previous models of the development of this margin, we show that at the time of formation of the basalts from Sites 989 and 990 (1) melting was at relatively shallow levels in a fully-fledged rift zone; (2) fragments of continental crust were present in the lithosphere above the zones of melt generation; and (3) the sublithospheric mantle was dominated by a depleted Icelandic plume component.

Description: © The Author(s), 2016. This article is distributed under the terms of the Creative Commons Attribution License. The definitive version was published in Geochimica et Cosmochimica Acta 176 (2016): 227–238, doi:10.1016/j.gca.2015.12.027.

Description: Measurements of Xe isotope ratios in ocean island basalts (OIB) suggest that Earth’s mantle accreted heterogeneously, and that compositional remnants of accretion are sampled by modern, high-3He/4He OIB associated with the Icelandic and Samoan plumes. If so, the high-3He/4He source may also have a distinct oxygen isotopic composition from the rest of the mantle. Here, we test if the major elements of the high-3He/4He source preserve any evidence of heterogeneous accretion using measurements of three oxygen isotopes on olivine from a variety of high-3He/4He OIB locations. To high precision, the Δ17O value of high-3He/4He olivines from Hawaii, Pitcairn, Baffin Island and Samoa, are indistinguishable from bulk mantle olivine (Δ17OBulk Mantle − Δ17OHigh 3He/4He olivine = −0.002 ± 0.004 (2 × SEM)‰). Thus, there is no resolvable oxygen isotope evidence for heterogeneous accretion in the high-3He/4He source. Modelling of mixing processes indicates that if an early-forming, oxygen-isotope distinct mantle did exist, either the anomaly was extremely small, or the anomaly was homogenised away by later mantle convection. The δ18O values of olivine with the highest 3He/4He ratios from a variety of OIB locations have a relatively uniform composition (∼5‰). This composition is intermediate to values associated with the depleted MORB mantle and the average mantle. Similarly, δ18O values of olivine from high-3He/4He OIB correlate with radiogenic isotope ratios of He, Sr, and Nd. Combined, this suggests that magmatic oxygen is sourced from the same mantle as other, more incompatible elements and that the intermediate δ18O value is a feature of the high-3He/4He mantle source. The processes responsible for the δ18O signature of high-3He/4He mantle are not certain, but δ18O–87Sr/86Sr correlations indicate that it may be connected to a predominance of a HIMU-like (high U/Pb) component or other moderate δ18O components recycled into the high-3He/4He source.