Fine-scale spatial heterogeneity of drought-induced mortality in tropical dry forests

Abstract:

Increasing extreme drought events have greatly enhanced global forest mortality risks in last decades. Multiple biotic and abiotic factors can interact to generate heterogenous mortality patterns during the same drought, especially in the diverse tropical forests. However, characterizing high-resolution mortality heterogeneity is still a challenge beyond field plot scale. Here, we integrated plot scale mortality observations and remotely sensed ecosystem resilience to generate fine resolution (i.e. 30m*30m) mortality maps for tropical dry forests in Costa Rica during the 2015 ENSO drought. Local water stress, forest composition, and topography all contributed to the spatial heterogeneity with varying relative importance across landscapes. Furthermore, we found that reduced ecosystem resilience can provide robust post-drought diagnosis of mortality but little early warning signals in our study region. These findings demonstrate fine-scale variations in ecosystem resistance to drought and the large potential of high resolution remote sensing time series to monitor forest mortality. Furthermore, we aim to apply our established methods to generate fine resolution mortality maps in Amazon forest, and we hope that we could tell this story by December.