For isotopic analyses, 5-cm3 samples from the carbonate-rich interval in Hole 1039B (Cores 170-1039B-21X through 41X) were examined at a ~10- (in Subunit U3B) to 5-m (in Subunit U3C) resolution. The same sample volumes were analyzed every ~8 (in Subunit U3B) to 4 m (in Subunit U3C) from the corresponding cores in Hole 1040C (Cores 170-1040C-36R through 52R). In addition to samples from Leg 170, a few samples from the lowermost interval of Subunit U3C recovered during Leg 205 were examined (5-cm3 samples from squeeze cake material from Cores 205-1253A-1R through 4R and 10-R through 12R; freeze dried at –80°C).
The oxygen and carbon isotope analyses of bulk sediment samples were conducted on a VG Isotech Prism Series II isotope ratio mass spectrometer fitted with a common acid bath automated carbonate device. Clean bulk samples (without pretreatment) were reacted in orthophosphoric acid maintained at 90°C by a water bath and cryogenically purified from water and noncondensable gases before introduction into the mass spectrometer. Analytical reproducibility is better than ±0.1
for 
13C and 18O values and was monitored through multiple analyses of a laboratory standard, which has been calibrated to the isotopic reference material National Bureau of Standards (NBS)-18 and NBS-19 for conversion to Vienna Peedee belemnite (VPDB) scale. Carbon and oxygen isotope ratios are reported in the conventional delta notation relative to the Vienna Peedee belemnite standard.
Sediment horizon correlations were used to compare the isotopic signals from Site 1039 with the signals from Site 1040. These were established by calculating the solid height profiles (i.e., the theoretical depth profile in a decompacted sedimentary section of solid material only, calculated from measured porosities data) at each site and by assuming constant accumulation rates (D. Saffer, pers. comm., 2002). The sediment horizon correlations are consistent with the boundaries between lithologic units and allow transforming Site 1040 depths to corresponding depths in the reference Site 1039. The transformation is depth dependent and has an estimated maximal absolute error of ~3.5 m.
To construct the age model, the results from Site 1039 were compared visually with the published isotope data from Woodruff and Savin (1991) and John et al. (2003). Maximum (13C) and minimum (18O) peaks in the measured data were matched to CM 1–6 and 18O Events A–F, respectively (Woodruff and Savin, 1991).
