ALBERT

Description: Authigenic metals (uranium, cadmium, and molybdenum), organic carbon (OC) and total C37 alkenone (totC37) concentrations were measured for the last 350 kyr in core MD900963, located in the eastern equatorial Arabian Sea. Authigenic metal concentrations on a carbonate-free basis range between 1 and 17 ppm, 0.5 and 6 ppm, and 0.5 and 4 ppm for U, Cd, and Mo, respectively. The profiles are characterized by well-defined 23 kyr cycles between oxic and mildly suboxic conditions. The redox-sensitive metal profiles also follow variations in the concentrations of OC (0.2-0.9%) and alkenones (0.2-6.7 ppm). The coupled variations in inorganic and organic constituents are attributed to a 23-kyr cycle in primary production above site MD900963, as suggested by clear correlations with independent micropaleontologic proxies (primary productivity indices based on foraminifera and coccoliths and fragmentation of foraminiferal shells). The 23-kyr cycles do appear to be primarily driven by productivity rather than changes in bottom water oxygen. Comparison with other records indicates that if this interpretation is correct, productivity variations across much of the Indian Ocean have been dominated by precessional forcing, with high productivity in phase with low summer insolation in the Northern Hemisphere. This interpretation contrasts with the traditional attribution of enhanced productivity in the Indian Ocean with periods of high summer insolation.

Description: A procedure is presented to separate diatoms and radiolaria from marine sediments and from each other, to purify them of elements associated with other phases, and to dissolve them to determine their elemental composition. The cleaning procedure eliminates artifacts due to the presence of detrital clays and the high sorption capacity of hydrated silica. The concentration of trace elements (Al, Fe, Mg, and Ba) that we find in alkaline dissolutions of clean diatoms are at least an order of magnitude lower than previously reported. The overall long-term precision in the determination of Ge/Si in a sub-standard of clean diatoms is ±0.024 * 10**-6 (1 sigma). Ge/Si measured in diatoms and radiolaria from core tops indicates that high-latitude Holocene diatoms accurately record the present-day oceanic Ge/Si, while radiolarian ratios are systematically lower and display more scatter. Evaluation of Ge/Si in diatoms and radiolaria from Hole DSDP 265 (Plio-Pleistocene) suggests that post-depositional alteration of the ratio does not occur at this site, but the average ratio carried by diatoms over this time interval was lower than that in the present ocean.

Description: We have determined germanium/silicon ratios in purified diatoms and radiolarians from siliceous sediments in Holocene core tops, one late Pleistocene piston core, and four high-latitude Southern Ocean Deep Sea Drilling Project (DSDP) sites ranging in age from Holocene to Oligocene. High-latitude Holocene diatoms in the 10- to 38-µm size fraction ([Ge/Si]opal = 0.693 +/- 0.039 * 10**-6) faithfully record the present oceanic ratio ([Ge/Si]seawater = 0.699 +/- 0.004 * 10**-6). This confirms that diatoms from siliceous oozes are monitors of seawater Ge/Si and may record past changes in rates of delivery of weathering products to the sea from continental and seafloor weathering. Large diatoms and radiolarians display ratios much lower than (Ge/Si)seawater, suggesting that (Ge/Si)opal, to some extent, displays the effects of biological fractionation. From 0 to 2.5 m.y.B.P., (Ge/Si)opal in diatoms from DSDP sites ranges from about 0.54 to 0.70 * 10**-6, similar to the range observed in one late Pleistocene piston core over the last 25 kyr. From 2.5 to 6 m.y.B.P., the ratio is more variable, while prior to 10 m.y.B.P., ratios are clearly higher, averaging about 0.9 * 10**-6 in the early to middle Miocene. Low values of the ratio are consistent with global weathering regimes dominated by river silica input to the sea, while higher ratios suggest periods of enhanced hydrothermal input or reduced fluvial contribution. Nevertheless, an oceanic one-box model with variable river and hydrothermal inputs and biogenic opal output cannot adequately account for the magnitude or the timing of changes observed in the record, suggesting the existence of other sources/sinks for inorganic germanium and/or silicon to or from the sea.