ARC Research Project

Has it always burned so hot? Fuel and fire changes in southeast Australian forests

An Australian Research Council funded Indigenous Discovery Project (IN21010005)

This project brings together classical Palaeoecology and cutting edge modelling and scientific methods. We will employ 4 key methodologies to address our 4 research questions:

 

Project summary

Indigenous cultural burning has been raised as a way of mitigating against climate-driven catastrophic bushfires in southeast Australian forests. It is argued that returning an Indigenous style fire regime will keep landscape fuel loads low, thus reducing the frequency and intensity of bushfires and mitigating against large catastrophic bushfires.

To explore this further and advocate for policy change, this project aims to bring together the enormous reservoirs of traditional fire knowledge in Indigenous communities with data-driven science by empirically testing how fuel loads, fuel type, fire frequency and fire intensity have changed in southeast Australian forests during the transition from Indigenous to British management over the past 500 years

Cutting-edge pollen-vegetation modelling

QUESTION 1 – FUEL LOADS: Have fuel-loads increased since the removal of Aboriginal management?

APPROACH: In this project, we will apply well-founded methods of estimating biomass change from subfossil pollen stored in wetland sediments. This procedure involves detailed mapping of forest inventories surrounding target wetland sites using a distance-weighted approach that involves direct biomass measurements and remote sensing. These contemporary forest inventories are calibrated against pollen accumulation identified from a fine-scale analysis of wetland sediments that have a robust radiometric age-control (such as those developed from the analysis of a combination of isotopic Lead, Caesium, Radium and Carbon within the wetland sediments). This calibration data can then be used to hind-cast forest biomass (at the level of individual plant taxa or community groups).  The assembled research team, particularly CIs Mariani, Connor and Fletcher, have a well-developed skillset in developing and implementing calibrations between pollen and distance-weighted vegetation sampling in Tasmania, mainland Australia and elsewhere. This phase of the project will require intensive field sampling of forest inventories and collection of wetland sediments targeted within southeast Australian forests systems.

We will develop pollen accumulation rate-plant biomass calibrations for plant species/pollen types ranging between low biomass (e.g. grassland, low heath) through to high biomass (tall forest). The number of sites is based on the literature underpinning these methods and the experience within the research team. The calibration will then be applied to pollen data analysed from sediment cores across the southeast Australian forest region.

HYPOTHESIS:   If cultural burning maintained a lower landscape-scale fuel loads in southeast Australian forests, our biomass reconstructions will show a systematic increase in woody biomass following the British Invasion and subsequent removal of cultural burning.

Classical Palaeoecology

QUESTION 2 – FUEL TYPE: Has the composition of fuels changed since the cessation of Aboriginal cultural burning?

APPROACH: In this project, we will apply novel modelling techniques to quantify past vegetation (i.e. fuel) cover, similarly to previous applications by the research team (Mariani, Connor, Fletcher) in Tasmania.  Pollen is the key proxy to track past vegetation changes, but pollen spectra suffer from some important biases (taphonomy, pollen productivity, dispersal capability). Pollen records are often dominated by pollen from a few high pollen-producing plant taxa (e.g. wind-pollinated plants), which mask the less producing taxa (e.g. animal-pollinated plants). In effect, there is a non-linear relationship between pollen percentages and plant cover. Estimating past vegetation cover from sedimentary pollen composition then requires to correct for productivity and dispersal biases using empirical-based models. Such models for quantitative vegetation reconstruction (e.g. REVEALS) have yet been mostly applied in the Northern Hemisphere in the last 10 years and can now be used in Australia thanks to the work conducted by the research team. By using this approach on newly developed pollen records, we will be able to quantitatively assess past fuel composition changes across SE Australia.

We will develop land cover-pollen calibration data for the same pollen taxa/plant species targeted for the biomass calibration. The calibration will then be applied to pollen data analysed from 10sediment cores across the southeast Australian forest region.

HYPOTHESIS:   If cultural burning maintained a more open (i.e. low near surface and elevated fine fuel)landscape in SE Australia, our land-cover reconstructions will show an increase in tree cover following British Invasion after the removal of Aboriginal management.

Palaeofire analysis

QUESTION 3:    Have fires become more frequent since the removal of Aboriginal management?

APPROACH: Our approach to reconstructing fire frequency involves two complementary approaches.  The first is founded on the well-established technique of charcoal peak analysis, which has been used in hundreds of fire history studies from around the world.  Peak analysis is performed using CharAnalysis software, which applies a robust lowess smoother to charcoal accumulation rates and then uses statistical thresholds to isolate individual fire ‘peaks’ from the ‘background’ charcoal produced by biomass burned on a regional scale. As part of this project, we will improve peak detection techniques to incorporate counting errors for the first time.  This improved technique will be coded in R and released for use by other researchers. We will compare peak detection results to a second approach, regime-shift analysis or change-point analysis, a novel technique for detecting rapid shifts in fire regime from sedimentary charcoal records. Both approaches to fire frequency reconstruction will be compared to historically documented catastrophic fires to: 1) provide a robust test of the fidelity of charcoal records from different sedimentary contexts and 2) calibrate the sedimentary signal to fire characteristics such as spatial extent, severity and frequency. This will constitute the first test of charcoal record fidelity in the Australian context.

The fire frequency analysis will be applied to charcoal data analysed from sediment cores across the southeast Australian forest region.

HYPOTHESIS:   Catastrophic fires have increased in frequency since British invasion due rapid accumulation of flammable biomass and ground fires, characteristic of cultural burning, have decreased in frequency since the suppression of Aboriginal management.

Fourier Transformed Infrared Spectroscopy

QUESTION 4 – FIRE INTENSITY: Have fires become more intense since the removal of Aboriginal management?

APPROACH: We will apply a state-of-the-art technique to reconstruct past fire intensity from fossil charcoal. The temperature of burning organic material results in a unique combination of combustion products which is captured in the chemical signature of charcoal. Fourier Transformed Infrared Spectroscopy (FTIR) will be used to analyse chemical spectra in fossil charcoal as a proxy for past fire combustion temperatures (fire intensity). The FTIR spectral absorbance of fossil charcoal fragments (>125 μm) will be recorded and the statistical analysis of FTIR spectra will be analysed in R studio using the Mclust and Chemospec to identify significant clusters of chemical spectra. These data will be compared with the reference database using a best-fit modelling to estimate fire intensity of fossil charcoal fragments. PI Maezumi has helped developed the protocol for analysing and calibrating fire intensity from fossil charcoal in the Amazon. This phase of the project will require lab analysis time that will be conducted at the University of Amsterdam.

The fire intensity analysis will be applied to charcoal data analysed from sediment cores across the southeast Australian forest region.

HYPOTHESIS:   If cultural burning maintained a lower landscape-scale fuel loads in southeast Australian forests, fire intensity will show lower fire intensities associated with cultural burning prior to the British Invasion.