Variations of sedimentary Fe and Mn fractions under changing lake mixing regimes, oxygenation and land surface processes during Late-glacial and Holocene times.

Global spread of anoxia in aquatic ecosystems has become a major issue that may potentially worsen due to global warming. The reconstruction of long-term hypolimnetic anoxia records can be challenging due to lack of valid and easily measurable proxies. The sedimentary Mn/Fe ratio measured by X-ray fluorescence (XRF) is often used as a proxy for past lake redox conditions. Yet the interpretation of this ratio can be problematic when Fe and Mn accumulation is not solely redox driven. We used the varved sediments of Lake Moossee (Switzerland) to examine the partitioning of Fe and Mn in seven fractions by sequential extraction under various oxygen conditions over the last 15,000 years. We combined this data with XRF scans and an independent diagnostic proxy for anoxia given by a hyperspectral imaging (HSI)-inferred record of bacteriopheophytin, to validate the use of the XRF-Mn/Fe ratio as redox proxy. In the 15,000-year long record, Fe was bound to humins and amorphous, crystalline, sulfide and residual forms. Mn was mainly present in carbonate and amorphous forms. Higher erosion, prolonged anoxia, diagenesis and humic matter input affected Fe and Mn accumulation. Under holomixis the XRF-Mn/Fe ratio successfully reflected lake redox conditions. Periods with higher detrital Fe input obscured the applicability of the ratio. During phases of permanent anoxia, intensified early diagenetic processes trapped Mn in the sediments in carbonate, crystalline oxide and humic forms. Our study shows that the single use of the XRF-Mn/Fe ratio is often not conclusive for inferring past lake redox conditions. The application of the XRF-Mn/Fe as a proxy for anoxia requires taking into account the individual lake characteristics and changes in lake environmental conditions, which affect the accumulation of Fe and Mn in the sediments.