Algal lipid distributions and hydrogen isotope ratios reflect phytoplankton community dynamics in Rotsee (Switzerland)

Reconstructions of past changes in algal community composition provide important context for future alterations in biogeochemical cycling. However, many existing phytoplankton proxies are indicative of individual algal groups and are not fully representative of the whole community. Here, we evaluated hydrogen isotope ratios of algal lipids (δ2Hlipid) as a potential proxy for phytoplankton community composition. We sampled the water column of Rotsee, a small eutrophic lake in Switzerland, every second week from January 2019 to February 2020 and analyzed distributions and the relative offsets between δ2Hlipid values (ε2Lipid1/Lipid2) from short-chain fatty acids, phytosterols and phytol. Comparing these data with phytoplankton cell counts, we found ε2C16:0/sterol and ε2sterol/phytol values reflect shifts in the eukaryotic algal community. To assess whether the selected phytoplankton groups were the main sources of the selected lipids, we further modeled algal ε2Lipid1/Lipid2 values based on δ2HC16:0, δ2Hsterol and δ2Hphytol values from batch cultures of individual algal groups and their biovolume in Rotsee and evaluated the role of heterotrophy on ε2Lipid1/Lipid2 values using a model incorporating δ2HC16:0 and δ2Hsterol values from microzooplankton. Annually-integrated and amount-weighted ε2Lipid1/Lipid2 values measured in Rotsee were within 6 to 17 ‰ of the mean of modeled algal ε2Lipid1/Lipid2 values, demonstrating a strong link with the phytoplankton community composition, while ε2Lipid1/Lipid2 values including microzooplankton lipids had a larger offset. Additionally, cyanobacterial biovolume was positively correlated with the ratio of phytol and phytosterols (phytol:sterol ratio) as well as the ratio of unsaturated C18 and C16:0 fatty acids (C18:C16 ratio). Our results support the application of sedimentary ε2Lipid1/Lipid2 values in eutrophic lakes as a proxy for past phytoplankton community assemblages. Moreover, the calculation of sedimentary phytol:sterol and C18:C16 ratios provides an additional proxy for reconstructing cyanobacterial blooms.

Methods
Water and biomass samples were taken biweekly from the water column of Rotsee at 1 m depth and the chlorophyll maximum depth over a one-year period. For biomass samples, lake water was filtered through glass fiber filters. The chlorophyll maximum depth on each sampling date was determined based on the turbidity maximum measured by a multi-parameter CTD probe. The temperature at the two different sampling depths was estimated by the mean of temperatures measured by the CTD probe from 0.5 m above to 0.5 m below the respective depth. Cell counts of phytoplankton and microzooplankton species were conducted from water samples and the biovolume of individual taxonomic groups was calculated based on cell densities and the mean per-cell biovolume. δ2Hwater values of filtered water samples were measured by a high-temperature conversion/elemental analyzer (TC/EA) coupled with isotope ratio mass spectrometry (IRMS). Nutrient concentrations of filtered water samples were measured via spectrophotometry (total phosphorus) and by a Total Organic Carbon Analyzer with Total Nitrogen Unit (total nitrogen). Lipids were extracted from biomass filter samples, purified, derivatized (methylated or acetylated) and quantified by gas chromatography–flame ionization detection (GC-FID). Compound-specific δ2Hlipid values of short-chain fatty acids, sterols and phytol were measured by GC-IRMS. δ2Hlipid values were corrected for hydrogen added during derivatization and δ2Hsterol values were further corrected for biases related to peak dimensions.