Percent calcium carbonate was determined on 328 samples (Table T1). Two intervals were analyzed, one within lithologic Subunit IA between 14.52 and 41.13 meters below seafloor (mbsf) (25 cm sampling interval) and one within Subunit IB between 67.24 and 100.12 mbsf (15–30 cm interval). Sample size was 20 cm3, which was contained in two 10 cm3 plugs taken from a 4-cm interval of the working half of the cores. Only one 10-cm3 plug, however, was utilized for wt% CaCO3, resulting in slight uncertainty in the exact downcore depth (i.e., the 10-cm3 plug from Sample 194-1198A-3H-1, 0–4 cm, may actually be located between 0 and 2 cm or 2 and 4 cm. Therefore, downcore depth is reported as the center of the interval, and a 2-cm uncertainty is acknowledged (e.g., 14.52 ± 2 cm for the above-mentioned sample).
A UIC CM5014 coulometer equipped with a CM5130 acidification module at the University of South Florida College of Marine Science (USA) was used to determine wt% CaCO3. Bulk sediment was placed in a drying oven for at least 24 hr, removed, allowed to equilibrate for ~1 hr, and subsequently ground and homogenized with a mortar and pestle. Approximately 15 mg of the finely ground bulk powder was removed, weighed, and placed into a sample vial for analysis.
Weak perchloric acid was used to dissolve the CaCO3 within a sample, and the mass of inorganic carbon (IC) within the evolved CO2 gas was measured by the coulometer. A clean cell and new anode and cathode solutions were used for each run (~40 samples). A "blank" analysis and at least four CaCO3 standards of varying mass between ~10 and ~20 mg were analyzed to verify proper performance. The mass of IC within each standard was determined stoichiometrically (e.g., mass IC = total mass x 0.12) and correlated to the mass of IC as determined by the coulometer. The resulting regression equation was subsequently used to determine wt% CaCO3 within each sample (Fig. F1). An additional standard and duplicate sample were run after every 10 and 20 analyses, respectively. Duplicate samples yielded a mean reproducibility of ±2.5%.