87Sr/86Sr analyses were conducted on pore-water samples obtained from drill cores collected at Sites 1049, 1050, 1051, 1052, and 1053 during Leg 171B (Fig. F1). Total penetration at these sites was 191.9, 606.0, 644.6, 684.8, and 183.2 m, respectively. Our samples were typically spaced between 10 and 25 m apart along the cores.
Whole-round sections of sediment were cut from the cores shortly after recovery on deck and extruded from the core liners, and the exterior of the sediment section was cut away to minimize seawater contamination. Pore water from this material was extracted using a hydraulic sediment squeezer. Shipboard pore-water chemical measurements are reported in Norris, Kroon, Klaus, et al. (1998). Water samples were split on shipboard and stored in flame-sealed glass ampoules for later analysis.
Water samples originally stored in flame-sealed glass ampoules were transferred by pipette to Teflon vials and weighed to seven significant figures. After weighing, a 84Sr spike was added to all samples for strontium concentration measurement and evaporated to dryness on a hot plate.
Sediment squeeze cakes were prepared for analysis by scraping or cutting the surface with a razor blade to remove possible surface contamination. An 8- to 20-mg portion of sample was weighed, and a known amount of 84Sr spike was added. Approximately 1 mL of ultrapure 10% acetic acid was added to each sample in a Teflon vial. The sample was allowed to digest for 30 min in cold acid, and then centrifuged. The solution was removed by pipette and evaporated to dryness.
All samples were reacidified with 250 mL 5 N nitric acid, and the solution was loaded onto a preconditioned crown ether cation exchange resin (SrSpec). Columns were washed twice with 500 mL 5 N nitric acid, and strontium was collected with two separate aliquots of 500 mL Milli-Q water. Two mL of 1 N phosphoric acid was added, and the solution was dried on a hot plate.
Isotope ratio measurements were performed using a VG Sector 54 mass spectrometer in the Department of Geological Sciences at the University of North Carolina at Chapel Hill. Isotopic values have been normalized to 86Sr/88Sr = 0.1194 using exponential normalization. The results for three to six (generally five) SRM-987 standards run in the same turret with unknowns were used to adjust the measured 87Sr/86Sr values of the unknowns; if 0.710250 was accepted as the correct value for the standard, and if the measured 87Sr/86Sr value for standards in a turret averaged 0.710245, then 0.000005 was added to the 87Sr/86Sr ratio for each sample in the turret.
Strontium concentrations (Fig. F2A) were computed using standard mass spectrometric isotope dilution techniques using the 84Sr spike. Replicate analyses of standard samples indicate that strontium concentration measurements have an accuracy and precision better than 1% of the measured concentration.
Host sediment ages were estimated using shipboard and shore-based refined biostratigraphic markers (Norris, Kroon, Klaus, et al., 1998). Assigned ages and their errors for nannoplankton and planktonic foraminifers within intervals between obvious sedimentation hiatuses were combined and plotted as a function of depth in each sediment core. A linear regression was fit to each selected depth interval and r2 was generally >0.94. Predicted host sediment age was computed using the regression equation and depth in the sediment. Strontium isotope values for the host sediment were assigned using version 2 of the strontium isotope look-up table of Howarth and McArthur (1997).