Close Window

CD-05 Abstract

The Present and Future Effects of Ground Fires on Forest Carbon Stocks, Metabolism, Hydrology, and Economic Value in Amazonia and Cerrado

Heloisa S. Miranda — UnB - Universidade de BrasÝlia (SA-PI)
Daniel Nepstad — Woods Hole Research Center (US-PI)


Amaz˘nian forests are burning beneath the canopy. Each year, accidental ground fires

burn a forest area that is similar in size to the area that is cleared and burned, with

large but poorly understood impacts on carbon stocks, forest metabolism, forest hydrology

and the prospect of sustainable forest management. But these ground fires are not included

in deforestation estimates. Forest fires have the potential to initiate a positive

feedback with local climate, because fire induced reductions in forest evapotranspiration

may cause declines in local rainfall, provoking additional burning of forests which are,

themselves, more susceptible to fire each time they burn. These fires may be the first

step in a process of "savannization", in which recurrent burning favors the

replacement of fire-sensitive trees by fire-resistant grasses and thick-barked trees, a

process that has already transformed much of the woodland forests of the Brazilian

Cerrado, adjacent to Amaz˘nia, into grass-dominated savannas.

In the proposed research, we will contribute to the land-use, carbon stocks and surface

water chemistry themes of the LBA-ECO campaign by investigating forest fire. We will

estimate the contribution of anthropogenic forest fires in Amaz˘nia and the Cerrado to

fluxes of carbon and water, and to the economic value of these forests, through a

combination of field experiments and measurements, landholder surveys, and regional

modeling. The proposed research will provide several data sets for the broader LBA

campaign. We will use eddy correlation towers to measure net ecosystem fluxes of carbon

and water vapor in four Cerrado woodlands with different levels of fire frequency. We will

measure the soil water balance to 8 m depth at these Cerrado sites, in two mature forests

and one logged forest equipped with towers in Amaz˘nia, and in one logged forest that is

experimentally burned. These measurements will allow us to determine the soil layers from

which water is being extracted to supply evapotranspiration and to provide an independent

estimate of evapotranspiration that will permit interpretation and testing of the

tower-based estimates of ET. We will also quantify several components of net primary

productivity that must be measured to interpret tower-based measurements of net ecosystem

productivity, including fine and coarse root biomass (to 8 meters depth), fine and coarse

litterfall, and stemwood increment. Our study will provide land-use history maps for 300

geo-referenced Amaz˘nian properties, stratified by property size and region, including

the timing of forest clearing, forest logging, forest burning, and agricultural management

practices, and the economic costs and benefits of fire prevention and fire control

practices. Finally, we will develop a mechanistic, predictive model that will allow us to

estimate the area of Amaz˘nian forest that burns each year, and associated changes in

carbon stocks and water flux. This model will allow us to begin to examine the impacts of

climate change and land-use policy change on forest fire. Hence, the proposed study has

the potential to contribute directly to the LBA goal of promoting sustainable land-use


Summary of Measurements

  • NEP, ET: Cerrado sites (Aguas Emendadas, IBCE reserve)

  • Litterfall, root biomass, AG biomass, deep soil moisture (ET, drainage): Tapaj6ˇs,

    Cerrado Sites, Sao Gabriel de Cachoeira (?), (and, through other projects: Caxiua˝a, Rio

    Branco, Paragominas)

  • Flammability: Tapajˇs, (and, through other projects, Paragominas, Santana de Araguaia,

    Rio Branco)

  • Property surveys of fire history, economic aspects of fire prevention and fire damage:

    five study regions (TailÔndia (e. Para), Mato Grosso, Rondonia, Rio Branco, ??)

Related Projects

  • Predicting the fire susceptibility of Amaz˘nian forests. (Nepstad, F. Brown, A.

    Setzer). Development of a model for Brazilian Amaz˘nia of the rainfall regime at which

    primary forests become flammable. NASA/MTPE/Teco. 1/97 through 12/99.

  • The effects of rainfall exclusion on forest flammability, carbon stocks and water

    relations of Amaz˘nian forests. (Nepstad, Adriana Moreira, E. Davidson, C. Klink, Claudio

    Reis de Carvalho, Moacyr Bernadino Dias Filho). Experimental drying of a one-hectare

    forest plot and pasture plots in Flona Tapajos and Cerrado sites, NSF Ecosystems, 9/97

    through 8/'00.

  • Modelling the effects of seasonal and inter-annual drought on the flow of water, carbon

    and nitrogen through Amaz˘nian forests. (Nepstad, E. Davidson, C. Potter, E. Maklouf).

    Adapting CASA model to Amaz˘nia; incorporating deep soil water uptake. NSF/Teco.


Close Window