Tracing the inputs and fate of marine and terrigenous organic matter in Arctic Ocean sediments: A multivariate analysis of lipid biomarkers

An understanding of the carbon cycle within arctic sediments requires discrimination between the terrigenous and marine components of organic carbon, insight into the removal mechanisms for labile carbon during burial and appreciation of shelf-to-basin processes. Using a large data set of multiple molecular organic markers (alkanes, alkanols, sterols, saturated and unsaturated fatty acids, dicarboxylic acids), we apply (1) principal components analysis (PCA) to obtain a robust comparison of biomarker compositions in Arctic Ocean sediments, (2) geometric mean (GM) linear regression of the PCA variables to estimate the relative contributions of labile/marine and stable/terrigenous sources to each biomarker and (3) the slope of the GM regression of each biomarker with TOC to provide a novel measure of the removal rate of each biomarker relative to phytol. The PCAand TOC-based indices generally increase together: biomarkers with very high TOC-based removal rates such as the saturated and unsaturated n-alkanoic acids generally have a high labile/marine content from PCA, while the sterols have low removal rates, but exhibit a range of labile/marine content values and the n-alkanes and n-alkanols have low values for both. A dominant feature of all PCA models examined is a progressive decrease in the autochthonous/marine biomarkers with each increase in sediment core depth, which points to a universal diagenetic alteration of organic carbon with depth in the cores. The PCA model also displays a shelf to basin trend that is non-diagenetic and implies the ongoing (centuries or more) delivery of long-chain n-alkanes, n-alcohols and nalkanoic acids in a matrix that is pre-formed and well-preserved within the sediments. Terrigenous biomarker distributions within the PCA model suggest that atmospheric transport of plant waxes in aerosols and the water borne transport of very fine plant macerals likely have significant roles in the export of these vascular plant biomarkers to the basins. Biomarker ratios and profiles of the PCA-based labile/marine content with core depth indicate that the PCA model is more strongly influenced by the biomarker lability than the marine content, while increases in the marine content are largely responsible for the shifts in composition for near-surface core sections. r 2005 Elsevier Ltd. All rights reserved.