Major and trace element content of the 17 lithostratigraphic units recovered during DSDP Leg 49 (Table 1)

IPOD Leg 49 recovered basalts from 9 holes at 7 sites along 3 transects across the Mid-Atlantic Ridge: 63°N (Reykjanes), 45°N and 36°N (FAMOUS area). This has provided further information on the nature of mantle heterogeneity in the North Atlantic by enabling studies to be made of the variation of basalt composition with depth and with time near critical areas (Iceland and the Azores) where deep mantle plumes are thought to exist. Over 150 samples have been analysed for up to 40 major and trace elements and the results used to place constraints on the petrogenesis of the erupted basalts and hence on the geochemical nature of their source regions. It is apparent that few of the recovered basalts have the geochemical characteristics of typical "depleted" midocean ridge basalts (MORB). An unusually wide range of basalt compositions may be erupted at a single site: the range of rare earth patterns within the short section cored at Site 413, for instance, encompasses the total variation of REE patterns previously reported from the FAMOUS area. Nevertheless it is possible to account for most of the compositional variation at a single site by partial melting processes (including dynamic melting) and fractional crystallization. Partial melting mechanisms seem to be the dominant processes relating basalt compositions, particularly at 36°N and 45°N, suggesting that long-lived sub-axial magma chambers may not be a consistent feature of the slow-spreading Mid-Atlantic Ridge. Comparisons of basalts erupted at the same ridge segment for periods of the order of 35 m.y. (now lying along the same mantle flow line) do show some significant inter-site differences in Rb/Sr, Ce/Yb, 87Sr/86Sr, etc., which cannot be accounted for by fractionation mechanisms and which must reflect heterogeneities in the mantle source. However when hygromagmatophile (HYG) trace element levels and ratios are considered, it is the constancy or consistency of these HYG ratios which is the more remarkable, implying that the mantle source feeding a particular ridge segment was uniform with respect to these elements for periods of the order of 35 m.y. and probably since the opening of the Atlantic. Yet these HYG element ratios at 63°N are very different from those at 45°N and 36°N and significantly different from the values at 22°N and in "MORB". The observed variations are difficult to reconcile with current concepts of mantle plumes and binary mixing models. The mantle is certainly heterogeneous, but there is not simply an "enriched" and a "depleted" source, but rather a range of sources heterogeneous on different scales for different elements - to an extent and volume depending on previous depletion/enrichment events. HYG element ratios offer the best method of defining compositionally different mantle segments since they are little modified by the fractionation processes associated with basalt generation.

Radiogenic isotope trace composition of the Labrador Sea

Geological reconstructions and general circulation models suggest that the onset of both Northern Hemisphere glaciation, 2.7 Myr ago, and convection of Labrador Sea Water (LSW) were caused by the closure of the Panama Gateway ~4.5 Myr ago. Time series data that have been obtained from studies of ferromanganese crusts from the northwestern Atlantic suggest that radiogenic isotopes of intermediate ocean residence time (Pb and Nd) can serve as suitable tracers to reconstruct these events. However, it has been unclear until now as to whether the changes that have been observed in isotope composition at this time are the result of increased thermohaline circulation or due to the effects of increased glacial weathering. In this paper we adopt a box model approach to demonstrate that the shifts in radiogenic isotope compositions are unlikely to be due to changes in convection in LSW but can be explained in terms of increases of erosion levels due to the glaciation of Greenland and Canada. Furthermore, we provide experimental evidence for the incongruent release of a labile fraction of strongly radiogenic Pb and nonradiogenic Nd from continental detritus eroding into the Labrador Sea. This can be attributed to the glacial weathering of old continents and accounts for the paradox that one of the areas of the world most deficient in radiogenic Pb should provide such a rich supply of radiogenic Pb to the oceans. An important general conclusion is that the compositions of radiogenic isotopes in seawater are not always a reflection of their continental sources. Perhaps more importantly, the transition from chemical weathering to mechanical erosion is likely to result in significant variations in radiogenic tracers in seawater.