High precision lead isotope systematics of lavas from the Hawaiian Scientific Drilling Project

We report Pb isotopic compositions for 35 samples of the volcanoes Mauna Loa and Mauna Kea from the Hawaiian Scientific Drilling Project (HSDP-1) core at Hilo. These data were obtained with an external precision of similar to 100 ppm (2 sigma(ext)) on the ratios (206)pb/(204)pb, (207)pb/(204)pb and (208)pb/(204)pb by Using a ph triple spike to correct for instrumental mass fractionation. The Pb isotopic compositions in the lower section (1200 to 280 m) of the con sample 200 to 400 ka-old Mauna Kea lavas, and display two well-defined linear arrays in (207)pb/(204)pb-(206)pb/(204)pb and (208)pb/(204)pb-(206)pb/(204)pb isotope spaces. There is a suggestion that Mauna Loa (0 to 280 m depth) also displays such linear array(s). However, analysis of the Mauna Loa samples is complicated by residual contamination and/or sample heterogeneity. While these latter data exhibit a satisfactory array in (208)pb/(204)pb (206)pb/Pb-204, there still remains scatter in (207)pb/(204)pb-(206)pb/Pb-204 space, making it difficult to assess the hue Pb isotope systematics of Mauna Loa. The presence of two linear Pb isotopic arrays in Mauna Kea can be interpreted as either reflecting two parallel isochrons or in terms of binary mixing. If interpreted as isochrons, the (207)pb/(204)pb-(206)pb/(204)pb systematics correspond to an age of similar to 1.9 Ga. Comparison of measured Th/U ratios in the lavas and those inferred from Pb isotope systematics strongly suggest that the Pb isotopic arrays reflect binary mixing, and this bears directly on the question of how many distinct components are present in the Hawaiian plume. Most of the new Mauna Kea data lie well outside the mixing-component triangle defined in the literature by the Kea, Loihi, and Koolau'' components. On the basis of the relationships between Pb isotope ratios in 3D and a principal component analysis of the Mauna Kea Pb isotope dataset, we show here that a three-component mixing model can in principle explain both mixing lines. However, such an explanation requires a highly specific set of mixing conditions in order to produce parallel arrays in Pb isotope space (2D and 3D). Therefore, our preferred interpretation is that the two arrays reflect binary mixing, with four discrete source components involved in the generation of the Kea lavas. Comparison of the Pb isotope characteristics of these lavas with those of East Pacific Rise (EPR) MORB glasses further suggests that EPR-type Pacific lithosphere does not contribute to the source of Kea lavas. The position of samples along the mixing lines does not correlate with stratigraphic height in the core, and therefore the age of the lavas. Rather, it appears as though the relative proportions of the endmembers are controlled by the spatial configuration of these endmembers, and by melting and transport processes in the source itself. The stratigraphic fluctuations of Pb and Sr isotopes contrast with the monotonic decrease of epsilon(Nd) and epsilon(Nf) values as a function of age. This may in part be explained by differences in analytical precision of isotope measurements relative to the total range of values observed. This analytical resolution is far higher for Pb than for the other radiogenic isotopes. Alternatively, the observed fluctuation may be caused by the mobility of lead (as well as Rb and/or Sr) during the ancient differentiation process that created the differences in parent-daughter ratios. (C) 2000 Elsevier Science B.V. All rights reserved.