Integrated Use of Experimental, Modeling and Remote Sensing Techniques to Investigate Carbon and Phosphorus Dynamics in the Humid Tropics

摘要

Moist tropical forests comprise one of the world's largest and most diverse biomes, and exchange more carbon, water, and energy with the atmosphere than any other ecosystem. In recent decades, tropical forests have also become one of the globe's most threatened biomes, subjected to exceptionally high rates of deforestation and land degradation. Thus, the importance of and threats to tropical forests are undeniable, yet our understanding of basic ecosystem processes in both intact and disturbed portions of the moist tropics remains poorer than for almost any other major biome. Our approach in this project was to take a multi-scale, multi-tool approach to address two different problems. First, we wanted to test if land-use driven changes in the cycles of probable limiting nutrients in forest systems were a key driver in the frequently observed pattern of declining pasture productivity and carbon stocks. Given the enormous complexity of land use change in the tropics, in which one finds a myriad of different land use types and intensities overlain on varying climates and soil types, we also wanted to see if new remote sensing techniques would allow some novel links between parameters which could be sensed remotely, and key biogeochemical variables which cannot. Second, we addressed to general questions about the role of tropical forests in the global carbon cycle. First, we used a new approach for quantifying and minimizing non-biological artifacts in the NOAA/NASA AVHRR Pathfinder time series of surface reflectance data so that we could address potential links between Amazonian forest dynamics and ENSO cycles. Second, we showed that the disequilibrium in C-13 exchanged between land and atmosphere following tropical deforestation probably has a significant impact on the use of 13-CO2 data to predict regional fluxes in the global carbon cycle.