Lipid compound classes display diverging hydrogen isotope responses in lakes along a nutrient gradient

Compound specific hydrogen isotope ratios (H-2/H-1) of lipid biomarkers preserved in sediments are used as paleohydrologic proxies. However, several variables, including contributions from different source organisms and their growth rates, can influence H-2/H-1 fractionation between lipids and source water. Significant uncertainties remain about how these factors combine to produce the net H-2/H-1 signal exported to sediments.To assess the influence of phosphorus availability on H-2/H-1 ratios of lipids accumulating in lake sediments, we analyzed surface sediments and sediment traps from ten central Swiss lakes representing a wide range of trophic states. In agreement with results from laboratory cultures, H-2/H-1 fractionation for the diatom biomarker brassicasterol (24-methyl cholest-5,22-dien-3 beta-ol) increased in more productive lakes (0.6 +/- 0.1 parts per thousand per mu g/L total P in sediment traps and surface sediments). In contrast, H-2/H-1 fractionation for phytol, the isoprenoid side-chain moiety of chlorophyll, decreased with increasing total P (-0.4 +/- 0.1 parts per thousand per mu g/L total P in sediment traps), suggesting that different biochemical mechanisms are responsible for changes in H-2/H-1 fractionation for each type of isoprenoidal lipid. Opposing changes in H-2-fractionation for sterols and phytol cause their H-2/H-1 ratios to converge as total P increases. This response may be a new tracer for phytoplankton growth conditions and is not influenced by the source water isotope value.Interpreting the H-2/H-1 ratios of short to long chain (C-14-C-30) n-alkanoic acids and n-alkanols is complicated by likely contributions from heterotrophs and/or vascular plants. These values generally did not correlate with lake water isotopes, nor did their fractionation factors correlate with total P. For most lipids there was no significant difference between sediment trap and surface sediment H-2/H-1 ratios. However, n-C-14-n-C-18 fatty acids were H-2-enriched in the surface sediments, most likely due to degradation in the water column. Our results indicate that interpretations of short-chain fatty acid H-2/H-1 ratios as a water isotope signal likely require supporting information about ecological conditions and community structure, but that paired H isotope measurements of phytoplankton-derived sterols and phytol may be developed as a proxy for phytoplankton growth. (C) 2018 Elsevier Ltd. All rights reserved