1993-1998: Amazonian Deforestation and Carbon Cycling in Oligotrophic Ecosystems: The Role of Heterogeneity, Succession and Patchiness (funded by NSF)

Description: Concern with the alarming rates of deforestation in the Amazon Basin has unwittingly led to overlooking the possible role of successional processes in tropical moist and rain forests. As much as 85% of areas cleared are recolonized by native tree species within a few years after clearing. Examination of these processes, as well as how local populations manage land in order to accelerate or retard succession, offers an opportunity to understand how such regions may be managed in the future, how carbon emissions at regional level may be stabilized, how climate and biodiversity are affected by differential successional rates, and how conservation and restoration projects may be implemented.

This project undertook a comparative analysis of deforestation and ecological succession in the Brazilian Amazon and the role of these processes in the global carbon cycle. Two oligotrophic sites were studied in the Zona Bragantina and one in the Rio Negro Basin. These three sites were contrasted to two eutrophic sites (one in the Lower Xingu basin and the other at Marajo island) studied for the past two years with support from NSF. At each field site, the following data were collected: (1) history of land use, (2) forest stand structure (i.e. composition, relative abundance and spatial dispersion of trees) (3) carbon content and nutrient characterization of plants and soils, and (4) climatological data to assess changes in temperature, precipitation, albedo and evaporation at sites with contrasting areal extent of deforestation. These field data were integrated with remotely-sensed imagery of these regions. An important component in this analysis was the use of multitemporal Landsat Thematic Mapper digital data to provide a regional-scale capability for assessing successional dynamics jointly with radar digital data to enhance the terrain analysis and GOES data to estimate volume and areal distribution of rainfall. A carbon balance model were developed using field data and geo-referenced to specific sites using GIS.

Current estimates tend to overlook the variability in rates of regrowth, and thus in differential carbon release and uptake in tropical moist and rain forests. We have found differences of up to a factor of seven in rates of regrowth as a function of differences in soil fertility, land use, and size of area cleared in the eutrophic areas studied. The procedures outlined here have proven feasible for the past two years' work in two eutrophic sites. It is necessary to extend the understanding gained at the eutrophic sites to the more common nutrient-poor regions of Amazonia if we are to understand the Amazon's contribution to global warming. To ensure the relevance of this study to issues of global warming and carbon cycling, the proposed project will have the collaboration of two climatologists, a plant ecologist, and a senior remote sensing specialist. During the 1992 field study, it came to our attention that a decline of 100 mm had taken place in the most deforested region studied, and that the number of rainless days had increased over the past 20 years. Given that both eutrophic and oligotrophic sites under study are largely in the eastern Amazon, this provides a unique opportunity to examine the possible role of changes in rainfall regime on the vegetation, on species diversity (with potential high mortality of endemics with narrow ranges of adaptability to drier conditions), and on land cover.