Publications
LCNR supported
Ecology
Remote Sensing
New Phytologist (2026)
- Tall trees (height ≥ 60 m) are keystone elements of tropical forests, strongly influencing biodiversity, carbon storage, and ecosystem resilience. Yet, their density and spatial distribution remain poorly quantified, especially in remote Amazonian regions, limiting our understanding of their ecological roles and contribution to forest–climate interactions.
- We combined airborne LiDAR data from 900 transects across the Brazilian Amazon with environmental predictors to model tall-tree density. Spatial extrapolations allowed us to generate regional distribution estimates and assess associations with climate, topography, and disturbance regimes.
- Our model predicts that tall trees are unevenly distributed, with c. 14% of the estimated density concentrated in c. 1% of the Amazon and c. 50% within c. 11%. The highest densities occur in Roraima and the Guiana Shield provinces, where water availability is high and lightning or storm incidence is low. Modeled density strongly correlates with aboveground biomass, highlighting the disproportionate contribution of tall trees to carbon stocks. We estimate c. 55.5 million tall trees across the Brazilian Amazon.
- These findings demonstrate that tall-tree distribution is a crucial but underused predictor for biomass models. Understanding their ecological and spatial dynamics is vital for forest conservation and climate-resilience strategies under increasing anthropogenic pressures.
Related Projects
Mapping the resilience of tropical forests and savannas to global environmental change
Climate change effect on tropical forests
Understanding nature recovery paths and ecosystem functioning through forests health assessments
Quantifying the health of forests ecosystems by means of earth observation can aid in understanding nature recovery paths and ecosystem functioning
Mapping nature recovery at scale
Our AI team is developing state-of-the-art AI approaches to combine different sources of data, including drones, satellite, survey data and social media, that are robust to a range of environmental scenarios, data noise and model reliability.
