Close Window

The contribution of fire to forest degradation in the upper Xingu basin

Douglas Morton, University of Maryland, (Presenting)
Ruth DeFries, University of Maryland,
Carlos Souza, Jr., IMAZON,
Andre Lima, INPE,
Guido van der Werf, Vrije Universiteit Amsterdam,

Fires that burn standing forests are a critical pathway of anthropogenic forest disturbance in Amazonia. Repeated exposure to fire may arrest normal forest succession; high canopy-tree mortality leads to an increase in liana, grass, and fern cover in a low-carbon forest structure maintained through frequent burning. Understorey forest fires are more extensive during drought years, although forests also burn during years with average and above-average precipitation. When does the first fire exposure occur, and what roles do the adjacent land use and prior disturbance history play in rendering forests susceptible to burning? We combine field measurements of burned forest, time series of MODIS and Landsat data, and Landsat-based deforestation and logging extent to characterize the patterns of forest disturbance from fire, logging, and deforestation in the upper Xingu basin. Time series of satellite data for burned forests show a distinct phenological profile of forest damage and recovery. In previous analyses of high-resolution imagery, canopy damage from fire was more frequently mistaken as deforestation than logging. Logging greatly increases the chances that a forest will burn; burned area often matches the spatial extent of logging operations at a time lag of at least one year, confirming previous research showing that unlogged forests are less flammable than logged forests under similar climatic conditions. Following an initial exposure to fire, forests may burn as frequently as every 2-3 years. Thus, the cumulative area of fire-related forest disturbance may increase slowly in non-drought years despite extensive forest burning.

Science Theme:  LC (Land Use and Land Cover Change)

Session:  3C: Land Use and Fire

Presentation Type:  Oral (view presentation (9770 KB))

Abstract ID: 48

Close Window