Intense mid-Holocene warming on highland Sumatra: insights from biomarker proxies

Despite immense progress in the last decades, the Holocene climate evolution is still poorly resolved, in particular in the tropics, and especially from a terrestrial perspective. Here we reconstruct 11 000 years of paleo-climate and environment in the western Indo-Pacific Warm Pool – “the heat and steam engine of the world” – by analysing biomarker and geochemical proxies in peat sediments from Sumatra.

We discuss the composition of archaeal and bacterial membrane lipids (branched and isoprenoid glycerol dialkyl glycerol tetraethers; GDGTs) and their relationship with temperature and other environmental conditions. By analysing the hydrogen isotopes of leaf waxes (dD_wax) we reconstruct past rainfall amounts.

X-Ray Fluorescence (XRF) derived geochemical composition reveals changes in erosional regimes. Additionally, we use long-chain n-alkane distributions, carbon and nitrogen analysis, and attenuated reflectance Fourier-Transform Infrared analysis (FTIR-ATR) to investigate changes in vegetation on the peatland.

Three main climate-environmental phases emerge in our record: 1) Relatively cold, dry and unstable conditions during the Early Holocene which is marked by high detrital input into the peatland. 2) Warm, wet and stable conditions coincided with the mid-Holocene period, 8.2 – 3.2 ka BP. Using a global peat-specific temperature calibration based on branched GDGTs (Naafs et al., 2017), we derive mean annual air temperatures peaking at 4.8 ka BP that are ~3 °C warmer compared to core-top and modern local weather station data. The warmest period is also the wettest according to dD_wax, which is further supported by GDGT and alkane distributions, and d13C values indicating aquatic biomass production. 3) At 3.2 ka BP, the climate abruptly deteriorated into colder and drier conditions and re-intensified erosion.

Surprisingly, slope wash events resulting in input of coarse detrital material into the core were most frequent during the dry periods. We suggest that this is related to a more variable hydroclimate with droughts and episodic heavy rains, likely associated with ENSO variability, causing increased erosion during the Early and Late Holocene.