Eocene sea temperatures for the mid-latitude southwest Pacific from Mg/Ca ratios in planktonic and benthic foraminifera

We have used laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) to measure elemental (Mg/Ca, Al/Ca, Mn/Ca, Zn/Ca, Sr/Ca, and Ba/Ca) ratios of 13 species of variably preserved early to middle Eocene planktonic and benthic foraminifera from New Zealand. The foraminifera were obtained from Ashley Mudstone, mid-Waipara River, South Island, which was deposited at bathyal depth (ca. 1000 m) on the northern margin of the east-facing Canterbury Basin at a paleo-latitude of ca. 55°S. LA-ICP-MS data yield trace element depth profiles through foraminifera test walls that can be used to identify and exclude zones of surficial contamination and infilling material resulting from diagenetic coatings, mineralisation and detrital sediment. Screened Mg/Ca ratios from 5 species of foraminifera are used to calculate sea temperatures from late Early to early Middle Eocene (ca. 51 to 46.5 Ma), a time interval that spans the termination of the Early Eocene Climatic Optimum (EECO). During this time, sea surface temperatures (SST) varied from 30 to 24 °C, and bottom water temperatures (BWT) from 21 to 14 °C. Comparison of Mg/Ca sea temperatures with published δ18O and TEX86 temperature data from the same samples (Hollis et al., 2009) shows close correspondence, indicating that LA-ICP-MS can provide reliable Mg/Ca sea temperatures even where foraminiferal test preservation is variable. Agreement between the three proxies also implies that Mg/Ca-temperature calibrations for modern planktonic and benthic foraminifera can generally be applied to Eocene species, although some species (e.g., V. marshalli) show significant calibration differences. The Mg/Ca ratio of the Eocene ocean is constrained by our data to be 35–50% lower than the modern ocean depending on which TEX86 – temperature calibration (Kim et al., 2008; Liu et al., 2009) – is used to compare with the Mg/Ca sea temperatures. Sea temperatures derived from δ18O analysis of foraminifera from Waipara show amplified variability relative to the Mg/Ca and TEX86 proxies. This amplified variability is probably a diagenetic effect although it is possible that this Eocene δ18O record contains an ice volume component — the amplification signalling that temperature changes may have been accompanied by growth and collapse of ephemeral polar ice sheets on timescales of ca. 0.5 Myr.

Research Highlights
► Temperatures extracted from variably preserved Eocene foraminifera by LA-ICP-MS.
► Mg/Ca temperatures validated by comparison with independent δ18O and TEX86 data.
► Eocene SSTs of 24 to 30 °C and BWTs of 14 to 21 °C for the mid-latitude south Pacific.
► Places constraints on Mg/Ca of Eocene seawater.