Exploring the Influence of Temperature on brGDGT Distributions in Chilean Lakes and Soils: A Comparative Analysis of In Situ Measured and Modeled Temperature Data

Branched glycerol dialkyl glycerol tetraethers (brGDGTs) are bacterial membrane-spanning lipids that change in response to temperature variations. Their adaptability to temperature and widespread presence in sedimentary archives make brGDGTs a valuable tool for reconstructing past temperatures. However, the models used to relate brGDGT distributions to temperature vary widely (e.g., global vs. regional vs. site-specific, differing subsets of brGDGTs, and brGDGT-based indices), leading to inconsistencies in their application and inaccurate temperature predictions in some locations. Using previously published lacustrine and soil brGDGT distributions, we determined whether the type of temperature data used for model calibration (i.e., in situ vs. modeled air temperatures) influences the degree to which temperature relates to brGDGT distributions and therefore the fidelity with which brGDGT-based indices (i.e., MBT′_5ME) may predict temperature. Accounting for differences in the number of samples among lake surface sediment and soil sample data sets, we find that the impact of temperature on brGDGT distribution significantly varies depending on both the temperature and sample data set used, with the most pronounced effects observed in lake surface sediment samples. Similarly, the MBT′_5ME index shows varied correlations with temperature across different temperature and sample data sets. These findings suggest that other factors influence brGDGT distributions to a greater degree than temperature in some locations and that these effects are obscured when samples are combined into global data sets. This insight helps to explain why global brGDGT-based calibration models may not accurately predict temperatures at specific locations and underscores the need for targeted (e.g., cluster-based, regional, or site-specific) temperature calibration models.