Management of fire-prone shrublands worldwide is often directed at reducing risk of fire through application of prescribed fire.  The rationale for such an approach is the widespread assumption that fire hazard is directly related to biomass (and so, time since previous fire).   While extreme weather events may override such management actions, the presumption is that lower fuel loads will lead to fewer wildfires under more moderate conditions, or more easily suppressed fires under a broader range of conditions.  However, this assumption is largely untested.  Working with land managers in Western Australia, we implemented a series of 18 experimental fires in calcareous Mediterranean-type shrublands with high species richness.  We evaluated whether fire behavior (rate of spread, intensity) was related to time since last fire and quantified how fuel profiles change with time since fire (range of time since fire 3-24 years).  We also compared observed fire behavior with that predicted by the pre-eminent fire prediction program used by managers, BEHAVE. 

Results/Conclusions

Live and dead biomass, litter continuity, and canopy continuity all increased with time since fire.  However, fire rate of spread and fireline intensity for planned fires under low to moderate fire danger weather did not correspond to time since fire.  Higher fire intensities were associated with greater spatial heterogeneity in temperatures.  Fire behavior predictions were consistently lower than field observations, suggesting a failure of model assumptions.  Thus, time since fire (increasing biomass) does not correlate well with fire hazard, suggesting that fuel management may not serve to appreciably alter fire spread, underscoring the importance of weather in driving fire behavior.