Tropical forests have evolved under perturbations from fire, drought, pest and disease outbreaks, and storm events, and are adapted to intermittent disturbance. However, the frequency and magnitude of disturbance events in the tropics has grown over the last several decades and is predicted to increase under a changing climate. There remains a great deal of uncertainty surrounding the response of tropical terrestrial carbon fluxes to shifting disturbance regimes and future climate change. Recent research (Liu et al. 2017) suggests that mechanistic differences may underpin carbon dioxide fluxes across the three tropical forest regions of South America, Africa, and Asia, as well as sub-regionally (Eldering et al. 2017). Heterogeneous responses of tropical ecosystems to extreme drought events reflect differences in climate forcing, edaphic conditions, and ecosystem histories (evolutionary, environmental, and anthropogenic).
Working in collaboration with the Asner Lab, we are integrating airborne remote sensing observations of forest structure and canopy trait information obtained from LiDAR and imaging spectroscopy, and ground-based observations, with the Ecosystem Demography 2 (ED2) terrestrial biosphere model, to investigate how and why forests differ in their sensitivity to disturbance. This research is focused on tropical forests in the Peruvian Amazon and Sabah, Malaysia, two high biodiversity regions in South America and Southeast Asia. While both regions exhibit high tree species diversity, forests in South America and Southeast Asia followed very different evolutionary trajectories, resulting in significantly different species composition and function. We are interested in: (1) How differences in species composition and structure interact with land-use and drought histories to affect ecosystem responses to current and future environmental variability; and (2) How does sub-regional and regional heterogeneity influence ecosystem resilience to climate change?
Carnegie Airborne Observatory (CAO) data in the Peruvian Amazon (a) showing forest functional diversity derived from imaging spectroscopy in (b) and (c). Different colors in (a-b) indicate varying combinations of remotely sensed canopy foliar nitrogen, phosphorus, and leaf mass per area (LMA), which can be used to identify functionally distinct species (c) (Asner et al. 2016). Figure adapted from (Asner et al. 2016).
Tree height, derived from CAO LiDAR data, in a lowland forest (a) and oil palm plantation (b) in Sabah, Malaysia.