Testing and improving brGMGTs-based paleotemperature estimates in peat sediments: Evidence from global surface peat samples and two well-dated Holocene peat cores in far Northwest China

Abstract

BrGMGTs (branched glycerol monoalkyl glycerol tetraethers) are proposed as potential proxies for temperature estimation due to their constrained and consistent biological origin. However, discrepancies observed between brGMGTs-based proxies and MAAT (mean annual air temperature) have raised concerns about their applicability. Here, we analyzed surface samples from three peatlands across China, and two well-dated Holocene peat cores. We found: 1) surface peat samples showed that %brGMGTs (abundance of brGMGTs relative to brGMGTs and brGDGTs; brGDGTs: branched glycerol dialkyl glycerol tetraethers) and H-MBT (methylation degree of brGMGTs) were ineffective to represent temperature. 2) Holocene records of %brGMGTs and H-MBT were uncorrelated with MBT’5ME records (previously established as reliable temperature records) in the same cores, as well as with regional other proxy temperature records. Our data, along with the global peat dataset, indicated that brGMGTs are ineffective for temperature estimation under cold conditions (MAAT <10 °C). This ineffectiveness may be due to unfavorable temperature niches, changes in aerobic/anaerobic conditions and microbial communities. By setting a threshold (%brGMGTs >4 %), we found that H-MBT showed a higher positive correlation with MAAT (R2 = 0.50, p < 0.01, n = 26), compared to all samples (R2 = 0.47, p < 0.01, n = 72). Although this improvement in correlation is modest, it helps exclude samples with MAAT <10 °C. Our revised method for temperature estimation using brGMGTs could potentially be applied to peat sediments with MAAT ranging from 10 °C to 27 °C, enhancing their application for paleotemperature reconstruction in future research.