METHODS

Sediment samples were obtained from Pleistocene sequences at nine of the thirteen sites occupied during Leg 175. Sediments from Sites 1075, 1076, and 1077 in the Lower Congo Basin record varying histories of fluvial inputs, seasonal coastal upwelling, and excursions of the South Equatorial Countercurrent. Sediments from Sites 1078, 1079, and 1080 on the Angola margin represent accumulation under low-productivity hemipelagic conditions that are influenced little by either fluvial contributions or coastal upwelling. Sediments from Site 1081 on the Walvis Ridge and Sites 1082 and 1084 in the Northern Cape Basin are influenced to different degrees by the elevated productivity associated with the Benguela Current. Samples were collected at a frequency of one per section meter at Sites 1075, 1076, 1077, 1080, and 1082 and one per core at Sites 1078, 1079, 1081, and 1084.

Samples were freeze-dried and then ground to a powder. Total carbon and organic carbon (TOC) concentrations were measured using a LECO CS-244 carbon/sulfur analyzer. Dried bulk sediment was first analyzed for total carbon and then digested with 2.4-N HCl to remove carbonates. The carbonate-free residue was washed thoroughly with distilled water, and then the residual carbon was measured. This value represents the TOC content of the sediment. The difference between the two carbon measurements gives the inorganic carbon content, which is reported as CaCO₃ concentration, assuming all of the inorganic carbon was present as calcite or aragonite. Precision of the analyses, based on duplicates of every tenth measurement and standards (pure calcium carbonate) for every fifth measurement, is ±0.10% and ±0.02% for total carbon and organic carbon, respectively.

Preparation of CO₂ gas for carbon isotopic analyses followed the routine procedures of Craig (1953). Carbonate-free sediment samples were combusted at ~850°C in the presence of Cu, CuO, and O₂ gas. Water was cryogenically removed by a dry ice and alcohol bath. The CO₂ was condensed using a liquid N₂ trap, and the noncondensable gases were pumped out. The volume of CO₂ was measured with a manometer that had been calibrated with a standard carbonate reagent. The CO₂ collected was then analyzed using a VG Sira 10 isotope ratio mass spectrometer at the National Sun Yat-Sen University, Taiwan. Results are reported in conventional values relative to the Peedee belemnite (PDB) standard. The analytical precision for the standards expressed as 1 is better than 0.06.

Biogenic opal concentrations were determined by a silica alkaline leaching procedure modified from Mortlock and Froelich (1989) by using different acid and base reagents. HCl was replaced by 0.5-M glacial acetic acid to prevent over-extraction of silicate (Murray et al., 1995). Instead of the 2-N Na₂CO₃ used by Mortlock and Froelich (1989), 0.5-N NaOH was employed as the basic leaching chemical for complete dissolution of opal. The values reported here as opal percent concentrations are calculated as opal = 2.4 × Si_opal (Mortlock and Froelich, 1989) using dry sediment weights. The overall relative precision of opal determination is better than 1.5%.