摘要：Precise descriptions of forest productivity, biomass, and structure are essential for understanding ecosystem responses to climatic and anthropogenic changes. However, relations between these components are complex, in particular for tropical forests. We developed an approach to simulate carbon dynamics in the Amazon rainforest including around 410 billion individual trees within 7.8 million km2. We integrated canopy height observations from space-borne LIDAR in order to quantify spatial variations in forest state and structure reflecting small-scale to large-scale natural and anthropogenic disturbances. Under current conditions, we identified the Amazon rainforest as a carbon sink, gaining 0.56 GtC per year. This carbon sink is driven by an estimated mean gross primary productivity (GPP) of 25.1 tC ha−1 a−1, and a mean woody aboveground net primary productivity (wANPP) of 4.2 tC ha−1 a−1. We found that successional states play an important role for the relations between productivity and biomass. Forests in early to intermediate successional states are the most productive, and woody above-ground carbon use efficiencies are non-linear. Simulated values can be compared to observed carbon fluxes at various spatial resolutions (>40 m). Notably, we found that our GPP corresponds to the values derived from MODIS. For NPP, spatial differences can be observed due to the consideration of forest successional states in our approach. We conclude that forest structure has a substantial impact on productivity and biomass. It is an essential factor that should be taken into account when estimating current carbon budgets or analyzing climate change scenarios for the Amazon rainforest.
