Carbon Dynamics in Vegetation and Soils along the Eastern LBA Transect
Plinio B. de Camargo, CENA/USP (SA-PI)
Susan E. Trumbore, University of California (US-PI)
Our overall goals are to quantify the contributions of different components of the carbon cycle to overall ecosystem carbon balance in Amazonian tropical forests and to undertake process studies at a number of sites along the eastern LBA transect to understand how and why these fluxes vary with site, season, and year. We are dividing this work into a number of specific tasks: (1) determining the average rate (and variability) of tree growth over the past 3 decades; (2) determining age demographics of tree populations, using radiocarbon to determine tree age; (3) assessing the rate of production and decomposition of dead wood debris; (4) determining turnover rates for organic matter in soils and the mean age of C respired from soil using radiocarbon measurements; and (5) comparing our results with models and constructing models to predict the potential of tropical forests to function as sources or sinks of C.
The overall picture of tropical forest C dynamics emerging from our Phase I studies suggests that the fraction of gross primary production allocated to growth in these forests is only 25-30%, as opposed to the 50% assumed by many ecosystem models. Consequent slow tree growth rates mean greater mean tree age for a given diameter, as reflected in our measurements and models of tree age. Radiocarbon measurements in leaf and root litter suggest that carbon stays in living tree biomass for several years up to a decade before being added to soils, where decomposition is rapid. The time lags predicted from 14C, when coupled with climate variation on similar time scales, can lead to significant interannual variation in net ecosystem C exchange.
In Phase II we will continue our measurements to observe how forest C budgets can vary from year to year with climate. In addition to filling gaps in our stand-level C budgets, activities we will (1) expand our efforts in integration across sites through modeling, and (2) augment our existing measurements where needed to facilitate modeling (through availability of continuous temperature and moisture data), or to fill gaps in coverage of C cycle components, and (3) continue measurements begun in the past year to study differences in C dynamics between primary forest and other land cover types (flooded forest and cerrado). Some more detailed physiological measurements begun by Chambers, Higuchi, Santos, and Tribuzy at the INPA ZF2 forest and FLONA seca floresta project will continue in Year 1.