Evaluation of Organic Geochemical Proxies in Sediments as Recorders of the History of Biogeochemical Dynamics in Lake Erie over the Last Century.

Abstract

Against a background of a well documented and reconstructed paleoenvironmental history of Lake Erie over the last century, the reliability and accuracy of commonly used organic geochemical proxies from sediments in paleoenvironment reconstructions were evaluated. Carbon cycling in Lake Erie is reflected by stable carbon isotopes of total organic carbon and calcium carbonate (δ13Corg and δ13Ccaco3), total organic carbon concentrations (TOC) and calcium carbonate concentrations (CaCO3). δ13Corg and TOC both record changes in lacustrine productivity and nutrient loadings reliably and directly. CaCO3 contents are primarily controlled by decomposition of TOC in the hypolimnion and sediments, reflecting the carbon cycling between the epilimnion and hypoliminion/sediment. The carbon cycling process indicated by δ13Ccaco3 variations is complicated and appears to relate to primary productivity, temperature variations, phosphorus distributions and/or the recycling of 12C in the epilimnion. Although stable nitrogen isotopes (δ15N) reflect nitrogen cycling, they also record multiple biogeochemical processes, including anthropogenic inputs, denitrification, primary productivity and establishment of invasive species at different stages of lacustrine history, rendering the interpretations of δ15N as paleoenvironmental proxies difficult. The concentrations of biomarker molecules, including hydrocarbons (HCs), fatty acids (FAs) and fatty alcohols (OLs), reliably record changes of lacustrine productivity while their molecular distributions reflect various aspects of paleoenvironmental conditions. The degree of post-burial diagenesis is closed linked with the abundance of branched and unsaturated FAs. When used together, carbon-chain lengths of FAs and OLs may reliably reflect the relative importance between terrigenous and aquatic sources for organic matter. Speculatively, uncommon HC distributions originate from microorganisms inhibiting an oil-contaminated environment. The effects of spatial and temporal variability of sediment cores on paleoenvironmental reconstructions are investigated by comparing four cores sampled at a common location but in different years, i.e., 1982, 1988, 1991 and 2003. The approximately consistent temporal trends of organic geochemical proxies in these four cores suggest both spatial and temporal variability are not significant enough in biasing long-term paleoenvironmental interpretations. Nevertheless, the differences among the cores in year-to-year proxy variations and proxy values, which are mainly accounted by spatial variability of sediment cores, indicate that high-resolution (i.e. annual) paleoenvironmental interpretations may not be reliable.