Deconvolving eastern Mediterranean planktic foraminiferal δ18O; a focus on sapropels and sea-level reconstruction

Abstract

Marine sediments from the Mediterranean Sea are ideal climatic archives. The semi-enclosed geography of the basin allows the climatic signals recorded in the sediments to be amplified, and high sedimentation rates enable high-temporal resolution of climatic reconstructions. These advantages have been exploited to develop a methodology to reconstruct sea level changes, a key parameter to understanding past climates. The Mediterranean sea-level method used a planktic foraminiferal stable oxygen isotope (δ18O) record to deliver the first millennially resolved sea-level reconstruction extending beyond 0.5 Myr, which is independent from deep-sea benthic δ18O. However, there are two main issues with the Mediterranean sea-level reconstruction, which I will address in this thesis.

First, the Mediterranean’s sedimentary record is punctuated by sapropels: periodic deep-sea anoxic events strongly associated with times of African monsoon intensification. The increased freshwater influx to the basin during sapropels decreases the δ18O of surface waters, creating anomalies in foraminiferal δ18O records which prevent sea-level reconstruction over these intervals. Despite extensive research on sapropels, the magnitude of monsoonal intensification and freshwater runoff, along with its influence on δ18O, remains elusive. To address this issue, I first present a suite of new palaeoenvironmental records from eastern Mediterranean sediment core ODP 967 for 1.4 to 0.6 Ma. The expression of sapropels deposited during different climatic conditions in a range of proxy records is used to characterise different ‘types’ of sapropel. Following this, a multi-site, multi-species dataset of planktic foraminiferal δ18O was compiled for a case study of last interglacial sapropel S5 (~128-121 ka). To quantitatively investigate the spatial variations in foraminiferal δ18O observed for the surface layers during S5, I develop a box model of the Mediterranean which is used in conjunction with the S5 dataset. This approach reveals the importance of surface stratification and a temperature concentration effect in producing light surface water δ18O anomalies during S5. The Mediterranean box model is then used to deconvolve the planktic δ18O signal, and to estimate the volume of African monsoonal runoff during S5. This is the first quantification of monsoonal freshwater runoff during a sapropel event. This method has potential to be applied to other sapropels, and ultimately help continue the Mediterranean sea-level reconstruction through sapropel intervals.

Second, the Mediterranean sea-level method assumes a linear relationship between sea surface temperature (SST) and global ice-volume through time. However, recent Mediterranean SST reconstructions for the past 3.5 Myr suggest that this assumption may not be valid. To tackle this issue, I use the available Mediterranean SST records to approximate changes in SST response to global ice volume for the past 2.6 Myr. This is then applied to recalculate the Mediterranean sea-level reconstruction. The new Mediterranean sea-level record has a lower mean sea level than the original, and is in better agreement with deep-sea benthic δ18O based sea-level reconstructions, in particular for interglacials.

This work advances our understanding of monsoonal freshwater inputs into the Mediterranean during a prominent sapropel event, and has improved the Mediterranean sea-level method to better constrain the sea-level record for the last 2.6 Myr. Show more