Methane Index and TEX86 values in cold seep sediments: Implications for paleo-environmental reconstructions

Abstract

The isoprenoid glycerol dialkyl glycerol tetraethers (GDGTs) derived from archaea are widely used in the reconstruction of past climate and environment through proxies such as the Methane Index (MI) and TEX86. A pre-requisite for the application of TEX86, which serves as a proxy for upper ocean temperature, is that the sedimentary GDGTs primarily originate from planktonic Marine Group I Thaumarchaeota. The MI is commonly used as a quality control measure for TEX86 reconstruction to identify samples affected by methanotrophic GDGTs. Recently, the MI has also been used for the reconstruction of past methane cycling. However, the spatial variability of GDGT-based proxies and the relationship between MI and TEX86 remain unclear. In this study, we generated a comprehensive suite of data, including gas, porewater, bulk sediment geochemistry, archaeal cell abundance, GDGTs and their derived proxies. We collected sediment cores from four study sites offshore Southwest Taiwan characterized by the absence or presence of anaerobic oxidation of methane (AOM) and differences in the depth of the sulfate-methane transition zone (SMTZ). The distance between the coring sites varied from similar to 20 cm to similar to 2 km. The geochemical and DNA data indicated the presence of SMTZs at depths ranging from 4 cm to 290 cm in sediments. AOM-related GDGTs were predominantly composed of GDGT-2, followed by GDGT-1, GDGT-0 and GDGT-3. Although MI values in the SMTZ could reach as high as 0.7, they did not strictly vary based on the SMTZ depth nor the methane consumption rate. This, coupled with the discrepancies in the downcore profiles of gas and porewater geochemistry compared to archaeal DNA and GDGTs, suggest that the duration of SMTZ presence could be a key factor influencing sedimentary MI values. We observed strong relationships between TEX86 and MI at AOM sites; however, the direction of these relationships varied across different locations. Despite this, TEX86-derived temperatures showed good agreement between sites (<1.5 degrees C between sites located similar to 2 km apart), and with climatology data. Consequently, our findings suggest that even in samples with high MI values above the threshold of 0.3, the bias on TEX86-derived temperatures may not be as significant as generally assumed.