The effectiveness of prescribed burning for protecting houses during wildfires in Australia

Abstract Background Approximately 14,000 houses have been destroyed during wildfires in Australia since 1900, and the number of houses lost per decade continues to rise. Greater application of prescribed burning is one strategy employed to mitigate this trend, but there is debate about its effectiveness. We sampled 1617 houses that were exposed to wildfire in southern Australia and explored associations between the proportion of these houses that were destroyed and the per cent of the landscape previously burnt (by prescribed burning or wildfire). Results After controlling for other variables that affected house loss (i.e. weather, terrain and woody vegetation near houses), the proportion of houses destroyed during wildfire decreased with the per cent of the landscape previously burnt in the upwind direction, but only where previous burning occurred within 5 years and < 3 km upwind from houses. However, we found the majority of sampled houses (69%) had no prescribed burning within 5 years and < 3 km upwind. The amount of woody vegetation within 40 m of houses had a stronger and more certain negative association with house losses than previous burning. Conclusions Our findings suggest prescribed burning: (a) is generally conducted too far from houses to reduce losses, unless it effectively halts wildfires before they reach houses; (b) could play a greater role in mitigating house losses during wildfires, but only if substantially more burning can be done within 3 km from houses; and (c) is less effective than removing woody vegetation immediately adjacent to houses (i.e. establishing defensible space). Overall, we found limited evidence that prescribed burning, as currently practised in Southern Australia, reduces house losses during wildfires.