METHODS

Interstitial water samples were squeezed from whole-round cores immediately after retrieval of the cores with the standard titanium-stainless steel ODP squeezer (Manheim and Sayles, 1974). Care was taken to minimize contamination from drill fluids by the removal of the outer sediment rim. The retrieved pore waters were subsequently analyzed on board the JOIDES Resolution for concentrations of major and minor constituents, including sulfate, sulfide, and chloride, using methods described by Gieskes et al. (1991) and D'Hondt, Jørgensen, Miller, et al. (2003). Pore water samples were immediately fixed after filtration with Zn acetate. After removing the ZnS by membrane filtration (0.45 µm), dissolved sulfate was precipitated quantitatively as BaSO₄ by the addition of a barium chloride solution. The precipitate was washed several times with deionized water, dried at 110°C, and baked (600°C) to remove residual organic matter and water. Sulfur isotope ratios (³⁴S/³²S) were analyzed by combustion isotope ratio monitoring mass spectrometry (Giesemann et al., 1994; Böttcher and Schnetger, 2004). After combustion of barium sulfate to SO₂ in an elemental analyzer (Eurovector), sample and calibration gas were introduced alternately in a He stream into a ThermoElectron Finnigan MAT Delta⁺ gas mass spectrometer through a ThermoElectron Finnigan MAT Conflo II interface. Stable isotope ratios ³⁴S/³²S are given in the -notation with respect to the SF₆-based Vienna-Canyon Diablo troilite (V-CDT) standard (Ding et al., 1999) according to the following equation:

³⁴S[] = (R_sample/R_V-CDT – 1) x 10³, (1)

where R = ³⁴S/³²S. Analytical stable isotope results are compiled in Table T2.