Isotope and trace element characteristics of a super-fast spreading ridge: East Pacific rise, 13–23°S

Abstract

Isotopic patterns of Nd, Sr, and Pb are remarkably coherent along the super-fast spreading portion of the East Pacific Rise from 13°S to 23°S. Between 15.8°S and 20.7°S, all three define a broad, smooth peak, which culminates at ∼ 17–17.5°S and is characterized by elevated 87 Sr 86 Sr , 206 Pb 204 Pb , and lower ϵ Nd (reaching values of 0.70271, 18.64, and +8.9, respectively). To the north and south this peak is flanked by ∼ 300 km long, isotopically homogeneous sections of ridge with higher ϵ Nd (∼ +10.9) and lower 87 Sr 86 Sr (∼ 0.7024) and 206 Pb 204 Pb (∼ 18.1). Although otherwise similar, these two sections differ from each other slightly in their 207 Pb 204 Pb and 87 Sr 86 Sr ratios. The isotopic peak corresponds to a region of greater axial cross-sectional area, but axial bathymetry and physical segmentation appear generally unrelated to mantle isotopic composition. However, an abrupt break in isotopic ratios does occur at the large, > 3 Ma, southward-propagating overlapping spreading center at 20.7°S, which marks the end of the south limb or flank of the isotopic peak. The peak itself appears to be a manifestation of large-scale binary mixing between material possessing at least mildly plume-like Nd, Pb, and Sr isotopic characteristics (most abundant at ∼ 17–17.5°S) and two slightly different high- δ Nd mantle end-members equivalent to those north of 15.8°S and south of 20.7°S. Helium isotopes also define a prominent along-axis peak, but it spans a much narrower range of latitude and is offset slightly to the north of those for Nd, Sr and Pb isotopes. The combined results suggest that a discrete mantle heterogeneity may be entering into the melt zone near 15.8°S and migrating southward as far as the 20.7°S overlapping spreading center. Isotopic variability at short length scales is very limited throughout the entire 13–23°S region. It cannot be solely a result of homogenization by petrogenetic processes, because there is a lack of corresponding uniformity in ratios of highly to moderately incompatible elements; also, isotopes do not correlate with major element indicators of degree of partial melting or differentiation, or, in general, with the secondary magmatic segmentation thought to reflect different partial melting domains. Therefore, the subaxial mantle must be isotopically well-mixed relative to the scale of melting. In part, this probably reflects: (1) a larger volume of melting per unit length of ridge; and (2) a greater flow of mantle into the subaxial melt zone at super-fast spreading; but also must represent (3) a reduced amount of real isotopic variability in the shallow asthenosphere, as emphasized by the regional isotopic uniformity north and south of the isotopic peak. Such large-scale homogeneity could be a result of enhanced convective asthenospheric mixing over a long period of time. It could also reflect a low, long-term input of continental, lithospheric, recycled slab, or plume-type material into the regional asthenosphere. Largely independent of the north-south isotopic patterns is a fairly regular, southward depletion in highly incompatible elements such as Rb and Nb, superimposed on which is sizable local variability. Because ratios of highly to moderately incompatible elements show little or no correlation with major-element indicators of degree of melting, much of the variation in highly incompatible elements must be caused by a different (probably larger) volume of mantle than that conferring the major element signatures, or by one (or more) event that preceded the main melting episode in the not too ancient past.