RESULTS

Conventional radiocarbon and calibrated ages are given in Table T1. Calibrations were carried out using the CALIB program (Stuiver and Reimer, 1993; v4.4.2) with the MARINE98 calibration curve. The local marine reservoir correction used was R = 238 ± 49 (Marine Reservoir Correction Database [MRCD; radiocarbon.pa.qub.ac.uk/marine/; Reimer and Reimer, 2001, supplementary material at www.calib.org] Site 108 [Taylor and Berger, 1967]) for Sites 1227–1231 and R = 58 ± 47 (MRCD Site 107) for Sites 1225 and 1226. We are aware that the R corrections of Taylor and Berger (1967) have been obtained from coastal shell samples and so probably represent the ¹⁴C reservoir effect of surface waters rather than that of the seabed. Given that deep waters usually have an older ¹⁴C age than surface waters, our use of this R correction should be considered as a minimum correction. Regardless of the age relative to present (AD 1950), application of a constant R downcore will allow us to calculate sediment accumulation rates that can be reliably compared, unless there have been significant fluctuations in atmospheric radiometric carbon production over the accumulation interval.

For all dates that fall within the marine calibration range (i.e., those less than ~21 ka), with the exceptions of Samples 201-1228B-1H-2, 92–93 cm (OZG107); 60–61 cm (OZH149); and 179–80 cm (OZH150), the age quoted in Table T1 is the median probability age, and the error is plus or minus one-half of the second standard deviation range, which contains 100% of the probability distribution. For the three exceptions listed above, the error is plus or minus one-half of the first standard deviation range containing the largest proportion of the probability distribution, these being 73%, 82%, and 52%, respectively. All ages and errors have been rounded according to the conventions of Stuiver and Polach (1977). With one exception (the reversal of ages for Samples 201-1227B-1H-1, 26–27 cm, and 54–56 cm, which we consider to be equivalent within the 1 error), all other ¹⁴C ages obtained for the Holocene and LGIT core sections are in correct stratigraphic order.

The radiocarbon ages obtained in this study allow division of the stratigraphic section at the Peru continental margin sites into an uppermost Holocene section underlain by an upper Pleistocene (MOI Stages 2 and 3) interval, where both are present. As described below, these can be equated with subtle changes in the lithostratigraphy. They reveal that the Holocene is probably not present at Site 1227 but is represented by >2 m of core at both Sites 1228 and 1229.