Abstract ID: 498

A Conceptual model of Ecophysiological Function across the Amazon Basin using a Synthesis of Observed and Model Data.

Observations of the seasonal cycle of Net Ecosystem Exchange of Carbon (NEE) are variable across the Amazon Basin. NEE has been observed to be nearly uniform through the year at sites such as Cueiras Reserve in Amazonas. A pattern of efflux during the wet season and uptake during seasonal drought has been observed at multiple sites in the Tapajos River National Forest, while the opposite pattern (uptake during wet season, efflux during drought) has been observed at several sites in Mato Grosso . Using a synthesis of eddy covariance flux observations from multiple sites across vegetation and moisture gradients in the Amazon basin and a biophysical model (Simple Biosphere Model, SiB), we construct a conceptual model of the patterns of ecophysiological function that determine annual cycles and interannual variability in observed NEE across tropical Amazonia. We identify several factors that determine the behavior of Gross Primary Productivity (GPP) and respiration fluxes, while NEE is determined by their sum. We evaluate our conceptual model across a pseudo-latitudinal gradient from extremely wet (> 2000 mm annual precipitation) in the north to drier cerrado (savanna) regions in the south. In the extremely wet regions near the equator we find little or no stress on photosynthesis and, subsequently, little or no variability in annual cycles of energy or carbon flux. In these regions NEE is determined by vegetation response to high-frequency meteorological events. As precipitation decreases, the flux seasonality increases and phase relationships between GPP and respiration determine annual NEE cycles. As GPP/respiration cycles transition from an out-of-phase relationship in wetter regions to an in-phase pattern in the cerrado, NEE changes from dry to wet season carbon uptake.

Session:  Carbon - Scaling carbon fluxes to the region from measurements in plots, towers, and aircraft.

Presentation Type:  Poster