ABSTRACT
The Amazonian forest is, due to its great size, carbon storage capacity and present-day variability in carbon uptake and release, an important component of the global carbon cycle. Paleo-environmental reconstruction is difficult for Amazonia due to the scarcity of primary palynological data and the mis-interpretation of some secondary data. Studies of lacustrine sediment records have shown that Amazonia has known periods in which the climate was drier than it is today. However, not all geomorphological features such as dunes, and slope erosion, which are thought to indicate rainforest regression, date from the time of the Late Glacial Maximum (LGM) and these features do not necessarily correspond to episodes of forest regression. There is also uncertainty concerning LGM carbon storage due to rainforest soils and biomass estimates. Soil carbon content may decrease moderately during the LGM, whereas rainforest biomass may change considerably in response to changes in the global environment. Biomass per unit area in Amazonia has probably been reduced by the cumulative effects of low CO2 concentration, a drier climate and lower temperatures. As few paleo-vegetation data are available, there is considerable uncertainty concerning the amount of carbon stored in Amazonia during the LGM, which may have corresponded to 44-94% of the carbon currently stored in biomass and soils.
