HR: 1340h
AN: B43A-0126    [Abstracts]
TI: Profiles of Trace Gas Concentrations in Undisturbed Forest in the Brazilian Amazon
AU: * Carmo, J
EM: jbcarmo@cena.usp.br
AF: Janaina Braga do Carmo, Av. Centen rio 303, Piracicaba, SP 13416-000
AU: Crill, P
EM: crill@geo.su.se
AF: Patrick Crill, Dep. Geology and Geochemistry Stockholm University, Stockholm, 10691 Sweden
AU: Dias, J
EM: jadson@cena.usp.br
AF: Janaina Braga do Carmo, Av. Centen rio 303, Piracicaba, SP 13416-000
AU: Camargo, P
EM: pcamargo@cena.usp.br
AF: Janaina Braga do Carmo, Av. Centen rio 303, Piracicaba, SP 13416-000
AU: Keller, M
EM: michael@kaos.sr.unh.edu
AF: Michael Keller, Complex Systems Research Center UNH, Durham, NH 03824 United States
AB: Globally, upland tropical forests are the largest natural source of nitrous oxide (N$_{2}$O). Soils of upland tropical forests generally consume methane (CH$_{4}$) although this process has only a minor effect on the atmospheric CH$_{4}$ budget. In this study, we investigate the concentrations of N$_{2}$O, CH$_{4}$, and carbon dioxide (CO$_{2}$) measured in profiles on towers in undisturbed forest at three Amazon forest sites located in the municipalities of Manaus, Amazonas, Melga‡o, Para (Caxiuana), and Sinop, Mato Grosso. We measured gas concentration profiles at six heights above the ground on during both wet and dry seasons in 2003 and 2004. Nylon tubes (0.95 cm OD) were installed on towers used for meteorological and flux measurements in LBA. Gas samples were drawn through teflon filters (1$\mu$m pore size) to a manifold and directed either to an infra-red gas analyzer (LiCor IRGA Model 6262), to sampling canisters, or to exhaust. During sampling periods, we maintained a continuous flow of at least 1 L min$^{-1}$ through all sampling tubes. CO$_{2}$ concentration data from the IRGA were recorded continuously using a Datastick analog to digital converter and a palm top computer. We removed air samples in electro-polished stainless steel canisters for off-site analysis of N$_{2}$O and CH$_{4}$ by ECD and FID gas chromatography respectively. Sampling times were selected based upon real-time measurements of CO$_{2}$ concentration. Relatively stable meteorological conditions at night led to consistent increases in CO$_{2}$ concentrations. At times we also observed increases in the concentrations of CH$_{4}$ and N$_{2}$O concentrations correlated with increasing CO$_{2}$. The source of the increasing CO$_{2}$ is most likely respiration by soil and above-ground organisms. Correlated increases in CH$_{4}$ and N$_{2}$O concentrations also likely result from biological activity in the soil and the canopy layer of the studied forests. Concentrations of these gases increase at night because the rate of gas emission in the canopy layer exceeds the rates of consumption, transport, and dilution. Fluxes of CH$_{4}$ and N$_{2}$O will be estimated as: Fx = ($\Delta$ [X] / $\Delta$ [CO$_{2}$]) * FCO$_{2}$. Where Fx is the ecosystem flux of the gas of interest, FCO$_{2}$ is the ecosystem flux of CO$_{2}$ and the ratio ($\Delta$ [X] / $\Delta$ [CO$_{2}$]) is determined from integrated profile concentrations. Micrometeorological and ecosystem studies in progress will be used to estimate ecosystem CO$_{2}$ flux. The preliminary results indicate that, as expected, the undisturbed forest ecosystem is a source of N$_{2}$O and that the ecosystem N$_{2}$O emissions are lower in the dry season than in the wet season. We found that the upland forest ecosystem appears to be producing CH$_{4}$ independently of season. The source of the CH$_{4}$ is unknown. Upland forest soils generally consume CH$_{4}$ throughout the year.
DE: 0400 Biogeosciences
SC: Biogeosciences [B]
MN: 2004 AGU Fall Meeting