Tracing the eutrophication history of Lake Baldegg using diatom-bound nitrogen isotopes

Diatom frustules are well-preserved in marine and lacustrine sediments over hundreds or even thousands of years. In addition, although only in very small amounts, they also contain organic matter within their siliceous structure. Previous applications have shown that the ¹⁵N/¹⁴N ratio of the organic nitrogen contained in diatom frustules (diatom-bound δ¹⁵N, or δ¹⁵N_DB) can be used as a proxy for nutrient cycling in the polar oceans, and that it is not affected by diagenetic effects. However, the applicability of this paleo-proxy to lacustrine sediments has never been tested. Here, we explore the use of δ¹⁵N_DB to reconstruct the history of nitrogen dynamics in Lake Baldegg (Switzerland) over the past 300 years. This lake was heavily eutrophied due to anthropogenic activities during the 20^th century, before the implementation of lake restoration measures (i.e., artificial aeration of the lake bottom since 1982). Using a multi-proxy approach (e.g., reflectance-inferred chlorophyll a and organic carbon accumulation rates, XRF sulfur counts, bulk isotopic composition, C:N ratio), we identified two distinct eutrophication phases (1880-1950 and 1950-1980) that were characterised by an increase in organic matter accumulation and primary productivity, the occurrence of bottom water anoxia, and a change in the origin of the bulk organic matter. The implementation of re-oligotrophication measures has led to the disappearance of anoxic conditions at the bottom of the lake after 1995, and a decrease in phosphorus concentrations in the lake (the latter observed in the monitoring data), which seems to have mitigated primary productivity and organic matter accumulation. δ¹⁵N_DB increased during the first phase of eutrophication, which could be due to extended denitrification in the water column in an expanding anoxic water column zone, and/or limiting N concentrations for phytoplankton growth, leading to increased nitrate utilization. During the second phase, δ¹⁵N_DB decreased, probably because fixed N in surface waters was no longer limiting for phytoplankton. After the implementation of re-oligotrophication measures, δ¹⁵N_DB increased again, possibly the isotopic imprint of external N inputs with a high δ¹⁵N signature, such as organic fertilizers (e.g. animal manure, compost). Additionally, the δ¹⁵N of hand-picked Daphnia ephippia are lower than, and show no consistent offset to, δ¹⁵N_DB, suggesting that the N isotope signal of δ¹⁵N_DB is not transferred to the upper trophic level in that lake. Finally, we measured the offset between δ¹⁵N_DB and δ¹⁵N_BULK providing insight into the effects of early diagenesis on the N isotopic composition of bulk sediments. In Lake Baldegg, the offset reversed after the lake was artificially oxygenated, indicating a role of sediment oxygenation in the diagenetic alteration on δ¹⁵N_BULK.