Evaluating the effect of prescribed burning on the reduction of wildfire extent in Portugal

Abstract

Prescribed burning (PB) is increasingly recognised as a viable, cost-effective technique for reducing wildfire risk. Yet, quantification of the effect of PB on the reduction of wildfire extent in southern Europe is non-existent. We used 35 years of fire mapping data in Portugal to analyse wildfire regime metrics in nine landscapes before (1985–2004) and during (2005–2020), the period when systematic use of PB to reduce fuel load began. We employed robust linear mixed models and generalized linear mixed models to examine the role of annual wildfire extent drivers at landscape scales and quantified the effect of PB on the reduction of subsequent wildfire extent, i.e., leverage. We found evidence of larger wildfires during the PB era, suggesting that other influences are counteracting its putative effect. The most important variables explaining wildfire extent were fire weather and ignition density, with positive effect, and past wildfire and PB extent, with negative effect, corresponding to leverage. We obtained a total past fire (PB and wildfire) leverage of 0.28, i.e., approximately 4 ha of past fire decrease subsequent wildfire by 1 ha, increasing to 0.84 if only landscapes with high wildfire incidence are considered. The analyses failed to unequivocally identify a direct leverage effect of PB but a 1:5 return-for-effort is reasonable as a conservative estimate. Increasing the size of PB treatments and treatment effort in the landscape in combination with fire-modelling assisted planning would enhance PB leverage, making it more effective at reducing wildfire extent, rather than functioning only as a supplement to the effect of previous wildfires.