Enhanced post-wildfire vegetation recovery in prescribed-burnt Mediterranean shrubland: A regional assessment

Abstract

The primary objective of prescribed burning (PB) treatment is to promote increased ecosystem resistance to high fire severity and ecosystem adaptive resilience after unplanned wildfires under worsening climates. Yet, empirical evidence involving regional-scale assessments with enough spatial representativeness on how vegetation will respond to PB-wildfire interactions in Mediterranean shrublands is absent. This is particularly relevant to provide comprehensive insights on the effectiveness of PB policies and programs in southern European countries, where shrublands are the main PB target. Here, we investigated the PB effect on vegetation responses in the short-term (1–3 years) after subsequent wildfire encounters in Mediterranean shrublands at the regional scale across mainland Portugal. We leveraged geospatial cloud-computing platforms and remote sensing data (optical and synthetic aperture radar -SAR-) and techniques with physical basis to procure reliable proxies for vegetation fractional cover (FCOVER), vigor (normalized difference vegetation index; NDVI) and structural complexity (cross-polarization ratio of C-band SAR backscatter data; SAR-CR) at the landscape scale throughout the time series. The role of site productivity and topographic configuration variables as putative drivers of vegetation responses was also analyzed. We found that the mitigating effect of PB on wildfire severity translated into stronger vegetation responses in treated areas than in untreated counterparts after the first PB-wildfire encounter. Although FCOVER, NDVI and SAR-CR recovered gradually in both areas, the recovery rate was markedly higher in PB-treated areas 3-years after the PB-wildfire encounter than at earlier times in the time series. The effect of PB treatment was stronger for the FCOVER recovery response than for NDVI and SAR-CR. The effect of the PB treatment on FCOVER and NDVI responses was modulated by site environmental characteristics, such that their differences between PB-treated and untreated counterparts were maximal for both higher altitudes and warmer, drier locations in the landscape.