Influence of vegetation type on brGDGTs: Results from surface soils beneath sub-alpine forest and grassland under the same climatic conditions

Branched glycerol dialkyl glycerol tetraether lipids (brGDGTs) are increasingly used for terrestrial paleotemperature reconstruction. However, there can be significant offsets between the estimated temperatures based on brGDGT distributions in globally-distributed individual surface soils and the corresponding instrumental temperatures suggesting that additional environmental and/or biological controls could influence these distributions. We investigated the influences of seasonality and vegetation type on the brGDGT distributions by collecting surface soils beneath sub-alpine forest and grassland in September (warm month) and January (cold month), within an identical climatic background, in southern China. The absence of apparent seasonal changes in the soil brGDGT distributions between the warm and cold months indicates an annual or longer turnover time of soil brGDGTs at our study site. However, there are differences in the surface soil brGDGT distributions expressed as MBT'_5ME (the methylation index that is related to mean annual temperature, i.e., MAT) between forest and grassland. Specifically, the forest surface soil MBT'_5ME values were generally lower than those of grassland surface soils, which is consistent with the relatively higher summer grassland surface soil temperatures. This reveals a seasonal (summer) bias in the brGDGT distributions, and the dominant influence of temperature and secondary/indirect influence of vegetation type on brGDGT distributions. Finally, for brGDGT distributions in the 288 globally-distributed surface soils with known vegetation types, the root mean square error (RMSE) between calculated and measured MAT is slightly decreased when the vegetation types are taken into account, which further indicates a possible secondary/indirect influence of vegetation type on surface soil brGDGT distributions.