The North Atlantic margins (Fig. 1) provide an excellent and arguably unique area in which to investigate the structure of a large mantle plume soon after its initiation. Rifting effectively dissected the ancestral Iceland plume head at the time of break-up and produced thick piles of basaltic lavas forming seaward-dipping reflector sequences (SDRS). These basalts allow a direct comparison with the steady-state situation observed today down the Reykjanes Ridge. ODP Leg 152 sampled a transect across the SE Greenland SDRS at 63ºN, and investigated the timing and character of the transition from continental to oceanic magmatism (Larsen et al., 1994; Fitton et al., 1995). Leg 163 aimed to complete this transect and to provide a second transect at 66ºN. Other sections of North Atlantic SDRS have been sampled on Hatton Bank (DSDP Leg 81) and the Vøring Plateau (ODP Leg 104).
The source of most of the pre-break-up and break-up basalts at 63ºN on the SE Greenland margin (Site 917, Lower and Upper Series, respectively) was depleted N-MORB mantle, though a group of flows high in the Lower Series had an Icelandic mantle source (Fitton et al., 1996). All the post-break-up basalt lavas at 63ºN (Sites 915 and 918) had a depleted Icelandic mantle source (Fitton et al., 1996). Basalt from SDRS further from the axis of the ancestral plume (Hatton Bank) had an exclusively N-MORB source (Brodie and Fitton, 1996), while those closer to the axis (Vøring Plateau) had an Icelandic mantle source (Viereck et al., 1988). The head of the Iceland plume at the time of continental break-up was therefore zoned, with a core of Icelandic mantle surrounded by a carapace of anomalously hot but compositionally normal N-MORB mantle. The transition zone between the Icelandic core and the
N-MORB carapace was sampled along the 63ºN transect, situated a little over half way from the inferred plume axis to the outer limit of the SDRS, and appears to have been narrow. Therefore, if the plume head was disc-shaped, only the central 30% of its volume was made up of lower mantle, the outer 70% being made up of upper (MORB-source) mantle. This zoning is very similar in scale and character to that seen today along the Reykjanes Ridge. Here the chemical and isotopic effects of the plume (LREE enrichment (Schilling et al., 1983), 87Sr/86Sr (Hart, 1973)) extend from the centre of Iceland at 65ºN to about 60ºN on the ridge, whereas the thermal effects (water depth (Schilling et al., 1983) and major-element composition of basalt (Klein and Langmuir, 1987)) extend to 55ºN. Magmatism in the North Atlantic Province began at approximately 62 Ma, some
8 my before formation of the SDRS. Even this early magmatism preserves a suggestion of compositional zonation in the
ancestral plume. By the time of SDRS formation, a compositionally zoned plume was clearly established, and this zonation has persisted to the present day. The thick carapace of hot MORB mantle suggests that the plume originated at a thermal boundary layer at the base of the upper mantle and entrained a core of lower mantle. It is unlikely that a plume head from the core-mantle boundary (Griffiths and Campbell, 1990), impacting at about 62 Ma, could have heated sufficient MORB mantle to produce the observed thermal and compositional structure.
Brodie, J. & Fitton, J.G., Proc. ODP, Sci. Results 152, (1996, in press).
Duncan, R.A., J. Geophys. Res. 89, 9980 (1984).
Fitton, J.G. et al., J. Geol. Soc., 152, 985, (1995).
Fitton, J.G. et al., Proc. ODP, Sci. Results 152, (1996, in press).
Griffiths, R.W. & Campbell, I.H., Earth Planet. Sci. Lett. 99, 66 (1990).
Hart, S.R. et al., Nature Phys. Sci. 246, 104 (1973).
Klein, E.M. & Langmuir, C.H., J. Geophys. Res. 92, 8089 (1987).
Larsen, H.C. et al., Proc. ODP, Init. Repts. 152 (1994).
Schilling, J.-G. et al., Am. J. Sci. 283, 510 (1983).
Viereck, L. et al., Geol. Soc., London, Spec. Publ. 39, 69 (1988).
White, R. & McKenzie, D., J. Geophys. Res. 94, 7685 (1989).
Fig. 1: Map of the North Atlantic region restored to anomaly 24 (54 Ma) showing the distribution of onshore and offshore Tertiary igneous rocks, the location of DSDP and ODP drilling transects, and proposed locations (A (Duncan, 1984) and B (White and McKenzie)) of the Iceland plume. A 1000-km radius area centred on B represents the possible size of the plume head at the time of break-up.