Distribution of glycerol dialkyl glycerol tetraethers in soils from two environmental transects in the USA

Glycerol dialkyl glycerol tetraethers (GDGTs) of both archaeal and bacterial origin form the basis of new temperature proxies applicable to soil, and lake and marine sediments. In soil, branched GDGTs are prevalent and their abundance of methyl or cyclic groups has been calibrated to mean annual temperature (MAT) using MBT and CBT indices. However, soil pH is also known to be an important variable controlling the distribution of branched GDGTs. Precipitation amount helps control soil moisture, as well as pH, and soil moisture is a leading variable affecting microbial diversity and activity in soil. We have evaluated the distribution of GDGTs from two soil transects in the USA: a dry, western transect covering six western states and a wet, east coast transect from Maine to Georgia in order to assess the effect of precipitation on the distribution of soil GDGTs. Our results show distinctly different GDGT distributions across climatic regions, with dry western soils characterized predominantly by thaumarchaeotal isoprenoid (iso) GDGTs and as a consequence, low BIT index values (0.2–0.6) and moist-temperate, east coast transect soils expressing mostly branched (br) GDGTs and higher BIT values (0.9–1). The predominance of iso GDGTs in the western soils is related to the degree of aeration, which in turn is related to precipitation amount, and also to soil pH. We also observed a substantial increase in the offset between measured MAT and MBT/CBT-based MAT below an annual precipitation of 700–800 mm yr−1, implying an impact of precipitation amount on MBT/CBT-based temperature reconstruction. The data suggest that, while soil tetraethers work well as a temperature proxy in moist-temperate regimes, they do not produce reliable measurements of temperature in sediments sourced from areas with < 700–800 mm yr−1 precipitation. Moreover, erosion of soils with low BIT values into lacustrine or marginal marine environments will not be detected via the BIT index, which can potentially affect paleotemperature reconstruction from sediments, and so provide erroneous estimates of soil carbon delivery. BIT index values also show a correlation with precipitation amount. The abundance of iso GDGTs in western transect soils allowed calculation of TEX86 values, but no correlation was found between TEX86 calculated temperature and mean annual temperature.