Sediment biomarkers elucidate the Holocene ontogeny of a shallow lake.

We carried out geochemical analyses on a sediment core from Lake Harris, Florida (USA) to identify sources of organic matter to the sediment throughout the Holocene, and relate changes in those sources to shifts in past climate and environmental conditions. We hypothesized that the sources of organic matter changed in response to regional hydrologic shifts following de-glaciation, and to human population expansion in the state during the 20th century. Hydroclimate shifts in Florida were related to: 1) a steady rise in relative sea level and the fresh water table that began in the early Holocene, 2) wetland formation and expansion ca. 5,000 cal yrs BP, and 3) the onset of the modern El Niño (ENSO) cycle ~3,000 cal yrs BP. Stratigraphic changes in sediment variables from Lake Harris reflect each of these hydroclimate periods. Early in the Holocene, Lake Harris was a marsh-like system in a relatively dry, open-prairie environment. Organic sediments deposited at that time were derived largely from terrestrial sources, as inferred from high TOC/TN ratios, a dominance of longer-chain of n-alkanes (n-C29-31), relatively negative organic carbon isotope values (δ13CTOC), and low biogenic silica concentrations. In the middle Holocene, a positive shift in δ13CTOC coincided with the onset of wetter conditions in Florida. Submerged macrophyte biomarkers (n-C21-23) dominated, and during that period bulk organic carbon isotope values were most similar to δ13C values of mid-chain-length n-alkanes. In the late Holocene, δ13CTOC values declined, CaCO3 levels decreased to trace amounts, organic carbon concentrations increased and diatom biogenic silica concentrations increased from 10 to 120 mg g-1. Around 2,900 cal yrs BP, the effects of ENSO intensified and many Florida lakes deepened to their current limnetic state. Concentrations of algal and cyanobacterial biomarkers in the Lake Harris core increased by orders of magnitude after about AD 1940, in response to human-induced eutrophication, an inference supported by values of δ15N that fluctuate around zero.