a Dual Proxy of Oceanic Weathering Inputs: Rare Earth Element Patterns and nd Isotopes from Metalliferous Sediment Cores

Recent interest in reconstructing Nd isotopic distributions in the past ocean using Fe-Mn crusts, oxide coatings on sediment particles, foraminifera shells and fish teeth has raised questions about relative influence of water mass mixing vs. variations in weathering input. We have explored the utility of hydrothermal metalliferous sediments for generating high-resolution records of both Nd isotopes and rare earth element (REE) patterns in deepwater. We present a 135kyr record of high precision REE data, preliminary Nd isotope ratios, and δ¹⁸O from a 2.3 m core collected at 3180m at 11S on the East Pacific Rise (EPR). REEs are dominated (\>96%) by seawater REEs adsorbed to Fe particles in these metalliferous carbonates; the hydrothermal REE source itself is negligible. Downcore variations in Nd/Er (proxy for light/heavy pattern fractionation) are not large (10%) but are easily quantified by our ICP-MS method (precision 1%), and reproducibly correspond to δ¹⁸O shifts. The largest Nd/Er excusions occur during deglaciations, e.g. Nd/Er=5.1 to 4.4 mol/mol between glacial MIS 6 and interglacial MIS 5e. If these variations reflect composition of ambient deepwater, then the results are opposite those expected by reduced influence of Atlantic component water (high Nd/Er) during glacials. Prelimininary ɛNd values for samples taken at Nd/Er extrema in MIS 1, 2, 5e and 6 are nearly invarient at -3.5+/-0.4, in excellent agreement with modern water column ɛNd for Pacific deepwater at this depth. Thus, while paleo-ɛNd shows virtually no change over 135kyr, REE pattern does carry information, but not, we believe, about water mass mixing changes. We suggest instead that Nd/Er reflects climatically-influenced weathering inputs to the Pacific REE pool. We propose a sea-level-driven mechanism: glacial low sea-stand reduces the extent of estuarine environments in which the light-REE-depleted seawater pattern is imparted, causing 1) more direct input of unfractionated crustal REE sources, and 2) release of REEs during erosion of high Nd/Er shelf sediments. This mechanism, unlike tectonically-driven shifts in weathering sources, can occur on glacial/interglacial timescales, and is expected to result in negligible change in Nd isotopic input, consistent with observations. Alternative mechanisms could involve temporal variations in the flux and composition of water column particles. In a first application to longer timescales, a low-resolution Cenozoic record from DSDP Leg 92, shows much larger but smooth variation of 50% in Nd/Er between 28 and 1.5 Ma, with high values from 28 to 20 Ma, decreasing during the mid-Miocene to values similar to modern at <8 Ma. ɛNd varies from -3 (near present deepwater value) to ~-6, but does not co-vary simply with Nd/Er. The combination of Nd/Er and ɛNd constitutes a new tandem proxy that may reveal changes in the style and/or source of weathering inputs to the past ocean, not interpretable from ɛNd alone. The results may have implications for climate-linked changes in the oceanic supply of biogeochemically important elements such as Fe and P.