On the Mechanistic Connection of Forest Canopy Structure with Productivity and Demography in the Amazon

摘要

Canopy structure has long been thought to influence the productivity and ecological dynamics of tropical forests by altering the availability of light to leaves. Theories and methods that can connect detailed quantitative observations of canopy structure with forest dynamics, however, have been lacking. There is urgent need to resolve this uncertainty because human-caused climate change may alter canopy structure and function in the Amazon. This work addresses this problem by, first, developing methods based on LiDAR remote sensing of fine-scale structural variation to predict the spatial structure of leaf area and light in forest canopies of the central Amazon (Appendices B&C). I show that LiDAR-based leaf area and light estimates can be used to predict the productivity of tree size groups and one-hectare forest plots--as well as differences between 2 sites separated by 500km (App. B). Sites also differed in canopy structure and the distribution of tree frequencies over size (size or diameter distribution). A model based on tree architecture, however, was able to connect observed differences in canopy architecture with size distributions to predict plot and site differences (App. D). This model showed that tree architecture is plastic in different light environments. While plasticity may increase light absorption, the smallest size groups appeared light limited. Absorption over size groups in one site, but not the other, agreed with the hypothesis of energetic equivalence across size structure. Ultimately, the performance of individual trees of different sizes in different canopy environments links forest demography with canopy structure and ecosystem function--I present a study aimed at improving tests of individual level theories for the role of light dependence in tree growth (App. A). Together, this work quantitatively connects canopy structure with forest carbon dynamics and demographic structure and further develops LiDAR as premier tool for studying forest ecological dynamics. Assessing variation in biomass growth and demographic structure over tropical landscapes with remote sensing will improve understanding of ecosystem function and the role of the Amazon in global Carbon dynamics.