Contrasting fire damage and fire susceptibility between seasonally flooded forest and upland forest in the Central Amazon using portable profiling LiDAR

Abstract

Abstract Fire is an increasingly important agent of forest degradation in the Amazon, but little attention has been given to the susceptibility of seasonally flooded forests to fire. Satellite images suggest that forests flooded seasonally by nutrient-poor black waters are more susceptible to fire and may suffer greater fire damage than nearby upland forests. Reasons for this difference may include the presence of a root mat, more fine fuel as litter and a drier understory in the flooded forest. We investigated this difference in the field, hypothesizing that differences in the aboveground structure of the pre-burn forest can contribute to the difference in impacts of, and susceptibility to, fires. We employed a portable profiling LiDAR (PPL), first to compare damage between adjacent black water seasonally flooded and upland forests that were burned by the same fire event, and to then assess pre-fire canopy structure attributes known to affect fire susceptibility. For both assessments, we used PPL-derived metrics of leaf area and vertical and horizontal variation in the structure of vegetation in the canopy. Four years after the fire, the LiDAR metrics showed greater combined effects of high damage and slow recovery in the seasonally flooded forest; reduction of total Leaf Area Index (LAI) after burning was only 10% for upland forest but was 71% in the flood forest. Compared to unburned upland, the canopy of unburned flood forest had structural differences that increase susceptibility to fire, including drier microclimate. It had more gaps, a more open understory and a lower upper canopy. Small patches lacking canopy closure (LAI