Huguet, A; Wiesenberg, GLB; Gocke, M; Fosse, C; Derenne, S: Branched tetraether membrane lipids associated with rhizoliths in loess: rhizomicrobial overprinting of initial biomarker record, Organic Geochemistry, 43(1), 12-19 (2012) | |
Abstract: Loess–palaeosol sequences are considered important terrestrial archives for studying Quaternary climate. Recent studies have shown that calcified roots (rhizoliths) can be significantly younger than surrounding terrestrial sediment (loess) and that loess organic matter may be contaminated with root-derived carbon from post-sedimentary deep rooting plants. Branched glycerol dialkyl glycerol tetraethers (GDGTs) are membrane lipids produced by unknown bacteria in soil and are increasingly being used as palaeoclimate proxies. The aim of this work was to examine the impact of the temporal decoupling between rhizoliths and loess on the abundance and composition of branched GDGTs, and consequently on the application of branched GDGTs as proxies in loess deposits. Branched GDGTs in rhizoliths and the surrounding sediment from a loess–palaeosol sequence in Nussloch (SW Germany) were examined. Their composition varied between the rhizolith and loess, which may reflect the fact that branched GDGTs were biosynthesized at different times (late Pleistocene for loess and Holocene for rhizoliths) and therefore under different environmental conditions. A large difference (ca. 7 C) in branched GDGT-reconstructed palaeotemperature was observed between rhizoliths and loess. In addition, the reconstructed temperatures were the opposite from those expected from the assumed depositional ages of branched GDGTs. This might be explained by the post-sedimentary input of root-associated, microorganism-derived branched GDGTs to loess, which might have led to an overprint of the initial branched GDGT signal in the sediment. Our results suggest that post-sedimentary penetration of deep-rooting plants in loess might entail uncertainty regarding the application of branched GDGT palaeoproxies and that caution should be exercised in interpreting temperature estimates derived from branched GDGTs in loess–palaeosol sequences. This is of special importance for loess, where rhizoliths are abundant, and where non-calcified root remains of different age from that of the sediment are possible. In addition, the GDGT concentration in rhizoliths and surrounding loess supports the hypothesis of a heterotrophic lifestyle for branched GDGT-producing bacteria. |