Carbon Cycling and Organic Radiocarbon Reservoir Effect in a Meromictic Crater Lake (Lac Pavin, Puy-de-Dôme, France)

Abstract

Lac Pavin is a meromictic maar lake for which the interpretation of sediment radiocarbon dates is complicated by the existence of a largely undefined reservoir effect resulting from degradation of carbon stored in the bottom layer of the water column. A data set of the contemporary 14C distribution of dissolved and particulate organic pools in the water column is presented to address this issue. Dissolved inorganic carbon (DIC) and organic carbon (DOC), plankton, suspended particulate organic carbon (POCsusp), sinking POC (POCsink), and bottom sediment organic carbon (SOC) were analyzed. Present-day Δ14C values of DIC were measured ranging from −750′ in the monimolimnion to atmospheric values in lake surface waters and in spring inlet waters. A range of Δ14C values between −200 and −300′ was observed for superficial POCsusp, POCsink, SOC, and DOC. This relatively uniform 14C offset of the exported organic production from the surface waters to the bottom represents a contemporary reservoir effect of ∼2500 yr. Laminated buried sediment samples and terrestrial vegetal macro-remains were used to evaluate temporal reservoir effect variations since the formation of the crater lake (7 ka cal BP). Buried sediment layers presented a similar offset or showed larger differences between Δ14C values of bulk sediment and terrestrial plant remains (–400 to −500′). Furthermore, an almost 0-yr reservoir effect was inferred from the sediment layers deposited just above the volcanic bedrock at the early flooding of the crater, and increasing slightly within the first centuries of the lake's history. A second objective was to tentatively model a defined scenario of the cycling of carbon in the lake capable of predicting a modern reservoir effect. Alternative scenarios were then tested for which a larger contribution of deeper DIC would provide a model compatible with larger past reservoir effects. It is concluded that using Δ14C SOC variation in laminated lake sediments as a proxy of paleolimnological conditions may be valuable provided that more data on the dynamics of the 14C composition of plankton and more detailed sampling of laminated sediment layers are available.