Degree Name

Doctor of Philosophy


School of Biological Sciences


Climatic change will potentially have wide ranging consequences for the impact of fire regimes on soil carbon pools in fire prone forests. Shifts in fire regimes are likely to result both from changing weather conditions and anthropogenic responses. Global change may require increased use of prescribed burning by land managers to mitigate the potential losses from an increased likelihood of wildfires. Such a change in management may entail trade-offs in other management goals such as maintenance of carbon stocks. Knowledge of the role of fire regimes in determining soil C concentrations is vital for predicting the consequences of changes to climate and land management options, as each of these processes will alter fire regimes. A series of studies in SE Australian eucalypt forests was carried out to provide a basis for evaluating how stocks of soil C may respond to coupled changes in climate, fire regimes and management. Specifically, the following studies were done to investigate how soil C concentration responds to fire regimes, climate and other disturbances:

  1. Time since fire (TSF) following fires of different intensity i.e., prescribed fires (TSF up to 43 years) and wildfires (TSF up to 38 years);
  2. Fire frequency (one, two and four fires in 30 years with TSF fixed at about 10 years) and variation in climate based on mean annual temperature (MAT) and mean annual precipitation (MAP);
  3. The impacts of four fire severity combinations from two successive wildfires with TSF fixed. The severity classes were: low (fire confined to the understorey), and high (fire causing substantial crown damage) and the combinations were low low, low high, high low and high high.
  4. Effects of experimental combinations of timber harvesting and varying frequencies of low intensity prescribed fires.

In all four studies the total C concentration in the mineral soil was estimated as a % by weight (%CTot). Recalcitrant pyrogenic C (%RPC) was estimated from C surviving a dilute Nitric acid and Hydrogen Peroxide digestion for experiments 1, 2 and 3. Total soil N expressed as % by weight (%NTot) was also estimated for experiment 4. A positive effect of TSF on %CTot was apparent in sites last burnt by wildfire more than 20 years previously, contrasting to prescribed fire which showed no effect as TSF increased. Climate had a stronger effect on %CTot than either fire frequency or intensity combinations of successive fires, with higher %CTot evident under cool and wet conditions. Relatively small fire frequency effects were climate region dependent. %CTot was higher after successive fires of low then high intensity compared with other intensity combinations. RPC was effectively a constant fraction of the CTot, irrespective of TSF, fire frequency and fire intensity. This questions how persistent RPC is in the landscape. Neither, timber harvesting or increased frequency of low intensity fire significantly impacted %CTot. By contrast %NTot showed a significant decline with increasing fire frequency in sites without timber harvesting. The response of %NTot in harvested sites, as a function of increasing fire frequency, was non-linear. Sites experiencing one or two post harvest fires had the highest % NTot but this declined quickly with additional fires reaching a minimum for sites with 4 or more post harvest burns. C:N ratios responded strongly to very high frequencies of low intensity fire, reaching values around 60, which may be sufficient to retard ecosystem function through nutrient immobilisation.

Overall trends in C and N metrics were discussed in terms of theoretical interplay between net primary productivity, vegetation decomposition and soil respiration and the way they were affected by climate, fire and timber harvesting, although these parameters were not directly quantified.

It was concluded that soil C (either total or recalcitrant fractions) was unlikely to be strongly affected by wide variations in fire regimes in these forests. The divergent response to fire type (wild fire or prescribed fire) in %CTot was only apparent at TSF longer than the average fire return interval for this landscape (i.e., 15 to 20 years). Thus, any attempt to substantially increase C sequestration in soils would require long-term exclusion of fire, the feasibility of which is questionable. Conversely, increased application of prescribed fire or occurrence of wildfires is likely to have negligible impact on soil C stocks in these forests. Soil C may therefore be robust to differing management strategies and future changes to fire regimes driven by climate change. By contrast, there is potential for climate change to strongly affect soil C, with a decline likely under future warmer and drier conditions, though the latter is uncertain.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.