Intact polar lipids as biomarkers for nitrogen fixation and nitrification in European soils

Rising atmospheric CO₂ levels are enhancing primary production, which could result in higher carbon sequestration in biomass (C-fixation). This higher primary productivity, however, relies on a sustained supply of soil nutrients, particularly nitrogen (N). To understand whether climate change and rising CO₂ levels have an impact on soil nitrogen availability on centennial to millennial timescales, a geological perspective can be used. For instance, the glacial-interglacial warming, and Holocene warm periods provide valuable natural analogues for investigating these long-term interactions. However, currently no quantitative methods exist to reconstruct soil nitrogen availability through time. To address this, we propose to develop novel proxies based on lipid biomarkers for essential processes in the soil N-cycle.

In aquatic systems, microbial membrane lipids are well established as biomarkers for specific steps in the N-cycle: for instance, heterocyte glycolipids (HGs), produced by heterocytous cyanobacteria, indicate biological nitrogen fixation. Further, microbial intact polar lipids (IPL), isoprenoid glycerol dialkyl glycerol tetraethers (isoGDGTs) and bacteriohopanepolyols (BHPs) are promising proxies for archaeal ammonia oxidation and bacterial nitrite-oxidation, respectively. We propose to test for these N-cycle biomarkers in soils. A pilot sample set of European surface soils was selected, consisting of both N-limited (N-) dryland (n = 8) and N-replete (N+) grassland soils (n = 4). Lipid extracts were analysed by UHPLC-HRMS to generate a high-resolution dataset of the complete lipidome. As a first step in the proxy development, we here present the BHP composition and changes in their relative distribution between in N- and N+ soils.

A total of 47 different BHPs were tentatively identified. In all samples, hydroxy BHPs are predominant components (50-60%). BHtetrol (BHT) is most abundant and all samples contain BHpentol and BHhexol. Amino-BHPs are less abundant in N- soils compared to N+ (2-7%; 10%). For example, aminotriol BHP is relatively increased in N+ soils. A total of 22 nucleoside BHPs were identified with either an adenine (adenosylhopanes) or inosine (inosylhopanes) headgroup that differ in amount and position of methylation on the BHP core or in the headgroup structure. Adenosylhopanes are relatively more abundant in N- than N+ soils (N-: 40-45%, N+: 20%). Inosylhopanes are present at a lower abundance, with 0-5% in N- and up to 10% in N+ soils. Based on changes in their occurrence, four adenosylhopanes and two inosylhopanes are tentatively proposed as N-cycle biomarkers. Specifically, three adenosylhopanes (diMe-adenosylhopane, diMe-adenosylhopane-headgroup-Me, Me-adenosylhopane-headgroup-diMe) and one Me-inosylhopane are exclusively found in N- soils. Likewise, an early adenosylhopane-headgroup-Me and an inosylhopane-headgroup-diMe only occur in N+ soils.

These results highlight the potential N-cycle lipid biomarkers in soils. The occurrence of other potential biomarkers (isoGDGTs, HGs) for the N-cycle will be tested for on the same soils. Moreover, we will apply an untargeted approach via computational MS to comprehensively characterize the whole soil microbial lipidome and evaluate the suite of potential lipid biomarkers associated with nitrogen cycling.