Nutrient limitation of CO2 fertilization in the tropics represents the single greatest source of uncertainty in the carbon cycle over the next half century, and must be parameterized in climate models. This project brings together a group of experimental, theoretical and modeling scientists with the goal of improving key processes that determine spatial and temporal variability of terrestrial carbon exchange within the GFDL dynamic land model LM3V. The team will use the Barro Colorado Island (BCI) research forests of the Smithsonian Tropical Research Institute in Panama as a model system for lowland tropical forests. The first step will be to characterize and parameterize the processes that generate and maintain nutrient limitation across tropical ecosystems. Next, the team will parameterize the physiological and growth responses of plants to variations in nutrient supply rates, using field-based experiments and observations. Finally they will evaluate the emergence of feedbacks between the coupled physical climate and terrestrial nutrient models.