EVOLUTION OF SEAWATER SR-87/SR-86 OVER THE LAST 400-KA - THE ABSENCE OF GLACIAL INTERGLACIAL CYCLES

Fluctuations in the 87Sr 86Sr ratio of seawater accompanying glacial/interglacial climate changes have been reported in recent studies but remain highly controversial. To investigate these potential fluctuations we present very high precision (13 ppm) 87Sr 86Sr measurements on planktonic foraminifera from Indian Ocean and Pacific cores. 87Sr 86Sr ratios from three different foraminiferal species are indistinguishable from one another in both core-top and 50 ka samples, demonstrating that changes due to diagenesis or contamination do not influence the measurements. Average 87Sr 86Sr ratios for Pacific Ocean and Indian Ocean samples are also indistinguishable at three intervals (core-top, 50 ka and 300-370 ka), demonstrating that the oceans have remained well mixed with respect to Sr. Also, 87Sr 86Sr ratios are not affected by changes in the precleaning of samples. Measurements from Pacific core V28-238, that used in the study of Dia et al. [1], do not reproduce the cycles seen in the previous study and are statistically well explained by a linear increase in the seawater 87Sr 86Sr ratio. It seems likely that an analytical artefact caused the cycles in the previous study. Measurements from two key sections of core from ODP site 758, that used in the study of Clemens et al. [2], do not reproduce the largest shifts seen in the previous study, despite the considerably better precision reported here. The apparent cyclicity in the Clemens et al. [2] data is also suggested to be due to an unknown analytical artefact. The 48 measurements presented here are a statistically good fit to a straight line defined by Δ87Sr(ppm) = -0.0613 × age (ka). At the 13 ppm level of precision no evidence for a glacial/interglacial variation in the seawater 87Sr 86Sr ratio is seen. Glacial/interglacial variation may still exist but this data constraints its maximum amplitude to 6-9 ppm, which would correspond to changes in the riverine Sr flux of ≈ 30%. © 1994