Lipid biomarker temperature proxy responds to abrupt shift in the bacterial community composition in geothermally heated soils

• Published in: Organic Geochemistry
• Main Authors: De Jonge, Cindy, Radujković, Dajana, Sigurdsson, Bjarni D., Weedon, James T., Janssens, Ivan, Peterse, Francien
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2019
• Subjects: Bacterial community; Biomarker lipid; Soil brGDGTs; Iceland
• Online Access: https://research.vu.nl/en/publications/f62b0a35-b1e2-436c-873c-387b9095d816; https://doi.org/10.1016/j.orggeochem.2019.07.006; https://hdl.handle.net/1871.1/f62b0a35-b1e2-436c-873c-387b9095d816; https://research.vu.nl/ws/files/237461480/Lipid_biomarker_temperature_proxy_responds_to_abrupt_shift_in_the_bacterial_community_composition_in_geothermally_heated_soils.pdf; http://www.scopus.com/inward/record.url?scp=85072022584&partnerID=8YFLogxK; http://www.scopus.com/inward/citedby.url?scp=85072022584&partnerID=8YFLogxK

Specific soil bacterial membrane lipids, branched glycerol dialkyl glycerol tetraethers (brGDGTs), are used as an empirical proxy for past continental temperatures. Their response to temperature change is assumed to be linear, caused by physiological plasticity of their, still unknown, source organisms. A well-studied set of geothermally warmed soils (Iceland) shows that the brGDGT fingerprint only changes when the soil mean annual temperature is warmer than 14 °C. This sudden change in brGDGT distribution coincides with an abrupt shift in the bacterial community composition in the same soils. Determining which bacterial OTUs are indicative for each soil cluster shows that Acidobacteria are possible brGDGT producers, together with representatives from the Actinobacteria, Chloroflexi, Gemmationadetes and Proteobacteria. Projecting the lipid fingerprint of the cold and warm bacterial communities onto the global soil calibration dataset creates two distinct soil clusters, in which brGDGTs respond differently to temperature and, especially, soil pH. We show that, on a local scale, a community effect rather than physiological plasticity can be the primary driver of the brGDGT-based paleothermometer along large temperature gradients. This threshold response needs to be taken into account when interpreting brGDGT-based paleo-records.