Palaeoceanography of the holocene and late-glacial N.E. Atlantic: development and application of biomarker proxies of environmental change

Abstract

The aim of this thesis is to develop and apply novel climate proxies to understand the palaeoceanographic evolution of the N.E. Atlantic during the late-Glacial and Holocene. The proxies investigated are based on organic molecular compounds called lipid biomarkers and bulk organic matter properties. The primary focus is on long-chain alkenones, molecules which have been extensively used in mid and low latitude open oceans to reconstruct sea surface temperatures (SSTs) during the Quaternary. Thus, the relative abundance of some alkenones is related to the growth temperature of the algae at the time of the biosynthesis of these molecules (expressed in the U(^K)(_37) and U(^K)(_37)’ indices). In high latitudes and coastal environments, the temperature dependence of alkenones is controversial, and the potential environmental information from alkenones is not yet well understood. In such locations there is increasing abundance of the C(_37:4) alkenone (quantified as %C(_37:4)). The presence of this component has been related to changes in the relative budget of freshwater in the surface ocean. A central aim of this thesis is to carry out an empirical investigation to find out the key environmental factors that control %C(_37:4) to assess its potential as a palaeoceanographic proxy. Research was conducted in the Nordic Seas and N.W. Scotland using samples from the water column, surficial sediment and sediment cores. The research undertaken can be broken down in three main sections: Alkenone distributions in the Nordic Seas. The aim was to clarify and extend the application of alkenones as palaeoceanographic proxies in subpolar to polar environments. Samples of filtered sea surface POM were analysed and extremely high %C(_37:4) values (up to 77%) were measured in polar waters (up to 80% sea-ice cover). Values of %C(_37:4) across the Nordic Seas showed a strong association with water mass type. A combined data-set revealed a stronger correlation of %C(_37:4) to sea surface salinity' (SSS, R(^2) = 0.72) than to SST (R(^2) = 0.5). However, scatter was observed in the relationship of %C(_37:4) to SSS, preventing confirmation of %C(_37:4) as a palaeo-SSS proxy. Values of %C(_37:4) in sea surface POM were high compared to surficial sediments. We discount preferential degradation of the %C(_37:4) alkenone and invoke dilution of the %C(_37:4) signal in sea surface sediments by advected allochthonous matter to explain this. The POM filter data suggest that, overall, U(^K)(_37) is a more appropriate SST index for the Nordic Seas than U(^K)(_37) '. Examination of the scatter in the U(^K)(_37) ' versus SST relationship, shows that regions in the south of the Nordic Seas (including the Icelandic shelf) may yield reliable, alkenone based, palaeoceanographic reconstructions. Comparison of %C(_37:4) distributions with dinocyst proxies in a late Holocene core from the Barents Sea suggests %C(_37:4) may be a general marker for the influence of arctic/polar water in palaeoceanographic reconstructions. The palaeoceanography of the Icelandic shelf iox. the post-Glacial period (0-15 kyr BP) was reconstructed from alkenone indices measured in three cores collected N and W of Iceland. One of the cores, JR51-GC35, contained a continuous record of Holocene sedimentation spanning 0 - 10.1 kyr BP. Superimposed on a general Holocene cooling trend in core JR5I-GC35 were millennial scale oscillations of ~2 C. The timing of the oscillations was in close agreement with the variability in IRD records from the East Greenland shelf and the timing of glacier advances in northern Iceland. A comparison of the U(^K)(_37) -SST records from JR51-GC35 and a published core from the eastern Nordic Seas (MD952011) showed significant differences (superimposed on the general trend) in the timing of millennial scale climate events. This illustrates that Holocene climate evolution in the Nordic Seas was more complex than previously suggested, with significant climatic differences between the eastern and western Nordic Seas caused by the differential variability of the Irminger and Norwegian Currents with time.The potential application for reconstructing past sea-level changes in NW. Scotland of lipid biomarkers (alkenones, n-alkanes and chlorophyll derivatives), and bulk organic parameters (%TOC, C(_nrg)/N) was assessed by a survey of modern basins (at different stages of isolation from the sea) and fossil basins (with known sea-level histories). A logit regression analysis of all the sediment samples was employed to find which of the biomarkers or bulk organic measurements could reliably characterize the sediment samples in terms of a marine/brackish or isolated/lacustrine origin. The results suggested an excellent efficiency for the alkenone index %C(_37:4) at predicting the depositional origin of the sediments. This study suggests alkenones could be used as an indicator of sea-level change in fossil isolation basins.