for 
18O (±1
). The values are reported relative to Peedee belemnite, based on calibrations directly to National Bureau of Standards 19.
The Mg/Ca analyses for Site 1241 (5°50“N, 86°26“W) were performed on the planktonic foraminifer Globigerinoides sacculifer. G. sacculifer has a supposed habitat depth within the mixed layer of 30–50 m (Fairbanks et al., 1982; Curry et al., 1983). For each sample, 20–25 specimens from the 355- to 400-µm fraction were picked. Specimens with a saclike final chamber or visibly contaminated specimens were not selected for analysis. In case of insufficient numbers of specimens, the 250- to 355-µm fraction was used to provide additional material, from which as many as 35 specimens were picked. The bias introduced by extending the size fraction amounts to <0.5°C (Elderfield et al., 2002). The shallow water depth (2028 m) of the core in comparison with the depth of the east Pacific lysocline (3600 m) (Lyle et al., 1995) indicates that no significant dissolution has occurred. Fe/Ca and Mn/Ca analyses were performed to monitor contamination by clays or Mn carbonates and showed that the samples were not contaminated.
After gentle crushing, the samples were cleaned according to the cleaning protocol of Barker et al. (2003). To remove clays, the samples were rinsed 4–6 times with distilled deionized water and twice with methanol (suprapur) with an ultrasonic cleaning step (2–3 min) after each rinse. Samples from several intervals were subject to severe fragmentation during ultrasonic treatment. Therefore, the duration of ultrasonic treatment was reduced to a maximum of 1 min per treatment for these samples. Subsequently, samples were treated with a hot (97°C) oxidizing 1% NaOH/H2O2 solution (10 mL 0.1-N NaOH [analytical grade] + 100 µL 30% H2O2 [suprapur]) for 10 min to remove organic matter. Every 2.5 min the vials were rapped on the bench-top to release any gaseous build-up. After 5 min, the samples were placed in an ultrasonic bath for a few seconds in order to maintain contact between reagent and sample. This treatment was repeated after refreshing the oxidizing solution. Any remaining oxidizing solution was removed by three rinsing steps with distilled deionized water. After transferring the samples into clean vials, a weak acid leach with 250 µL of 0.001-M HNO3 (subboiled distilled) was applied with 30-s ultrasonic treatment and two subsequent rinses with distilled deionized water. After cleaning, the samples were dissolved in 0.075-M nitric acid (HNO3) (subboiled distilled) and diluted several times so that all samples were expected to have Ca concentrations in the range 30–70 ppm, the ideal range for analysis (Garbe-Schönberg et al., unpubl. data).
Analyses were performed on a simultaneous, radially viewing inductively coupled plasma–optical emission spectrometer (Ciros CCD SOP, Spectro A.I., Germany) at the Institute of Geosciences (Kiel University, Kiel, Germany). A cooled cyclonic spray chamber in combination with a microconcentric nebulizer (200 µL/min sample uptake) was optimized for best analytical precision and minimized uptake of sample solution. Automated sample introduction was via an autosampler (Spectro A.I.). For Ca, we used the spectral line with the highest stability (183.801 nm), providing a truly simultaneous analysis with Mg and Sr within the same detector read-out phase. For Mg and Sr, we used the emission lines with the highest signal-to-noise ratios (279.553 nm for Mg and 407.771 nm for Sr). For Fe and Mn, we used the emission lines with the highest sensitivity (259.940 nm for Fe and 257.610 nm for Mn). Matrix effects caused by varying concentrations of Ca were cautiously checked and were found not to be significant (Garbe-Schönberg et al., unpubl. data).
Analytical sessions with batches of 200–300 samples usually took between 20 and 30 hr, during which the drift of the instrument could be neglected, being <0.2% as determined by analyzing an internal consistency standard after every five samples. The analytical error for analysis of Mg/Ca ratios was 0.1% for a total of 600 samples. Replicate analyses on the same samples, which were cleaned and analyzed during different sessions, showed a standard deviation of 0.09 mmol/mol, introducing a temperature error of ~0.5°C. The conversion of foraminiferal Mg/Ca ratios into SSTs was carried out by applying the equation of Nürnberg et al. (2000):
For stable isotope analysis, 10 specimens of G. sacculifer (without saclike final chamber) were picked from the 355- to 400-µm fraction. All isotope analyses were run at IFM-GEOMAR (Kiel, Germany) on a Finnigan Delta-Plus Advantage mass spectrometer coupled to a Finnigan/Gas Bench II. Analytical external precision was better than 0.07 for 18O (±1). The values are reported relative to Peedee belemnite, based on calibrations directly to National Bureau of Standards 19.
The age model for Site 1241 was constructed by matching cyclic variation patterns in climate proxy records of 18Obenthic, 13Cbenthic, and percent sand with patterns of changes in solar radiation that are controlled by cyclic variations in Earth's orbital parameters. This astronomically derived age model is in agreement with the most recent astronomical polarity timescale and with other orbitally tuned age models. The establishment of the age model is described in detail in
Tiedemann et al. (this volume).
