Lay summary

Cs) and thereconstruction of the past vegetation (pollen). Quantifying only(microscopically) visible charcoal may reflect charcoal from forest fires whichare relatively large in size and structurally sound. However these techniquesare less likely to quantify smaller charcoal fractions derived from grasses -probably the main contributor of charcoal in Australia’s vast savannas and opengrassy woodlands.137Pb, 210C, 14The fire regimes ofAustralia, the most fire prone continent on earth, have been changing duringthe late Quaternary and up to the present under the influence of a changingclimate and vegetation, Aboriginal impact and then by European settlers.Because fire history is an important parameter in understandingpalaeoenvironmental conditions in many parts of the world, it has beenreconstructed primarily by palynologists using lake cores and traditional tools(visible charcoal), combined with dating (

Therefore, we aredeveloping a new methodology to infer past wildfires by using geochemical toolsthat potentially assess the whole range of fire residues in sedimentary recordsand that can yield additional information about the vegetation burned. Inparticular, we propose that a geochemical marker method (benzene polycarboxylicacids (BPCA)) would be capable to detect sedimentary fire residues that are toosmall to detect with standard microscopic methods. So far, however, thesegeochemical markers have not been used to quantify fire residues in lakesediment cores, neither have they been cross-compared to the presence ofvisible charcoal, which is indicative of palaeofires.

The proof-of-concept studyis conducted at two Australian sites where we would use molecular markers(BPCA) together with other geochemical methods to quantify past occurrences offire and burned vegetation types. First we screen samples from about 200 depthintervals with a relatively rapid technique (MIR-PLS, mid-infrared spectroscopywith partial least square analysis) to observe major organic and inorcanicproperties. Then, an in-depth, and more time-consuming characterization followson some 20 samples from those sections of the cores, which have been identifiedby MIR-PLS to show significant changes in charcoal and organic carbonabundance. These sections will be analyzed using more sophisticated molecularscale techniques including the BPCA molecular marker method.

To summarize, the project takes up three tasks. 1. We systematicallycompare available organic geochemical tools to the palynological methodologyand try to decipher more information about fire history out of sedimentaryrecords. 2. More concretely, we want to test the hypothesis whether Australia’sHolocene fire record is biased towards forest fires and consequentlyunderestimates the contribution of grassland fires to the fire record andAustralia’s charcoal pool. 3. The methodology used has the potential to becomea quantitative approach to reconstruct wildfires. We will therefore try tosimplify the procedure for a straightforward use in the palaeosciencecommunity, for which it could be a valuable tool for future palaeofirereconstructions.