Automated Chamber Measurements of Soil-Atmosphere Carbon Dioxide Flux in Undisturbed Forest at the Tapajos National Forest, Brazil

In forests, the respiration of roots and soil dwelling organisms accounts for a large part of ecosystem respiration. We installed an automated chamber system for measurement of the soil-atmosphere flux of carbon dioxide (CO₂) in the Tapajos National Forest, Para, Brazil in April 2001. This is a mature forest site that is relatively undisturbed. Soils are clay textured oxisols. Mean annual temperature is 25°C and mean annual precipitation is 2000 mm per year. Eighteen aluminum chambers were installed in a 0.5 ha area close to the flux tower at the km 67 LBA site. Vegetation was excluded from the chambers. Eight of these chambers are closed individually and sampled for approximately 21 minutes about 5 times per day (closed 7% of the day). The other 10 chambers are sampled individually approximately once per day (closed 1.5% of the day). We measured CO₂ concentration with an IRGA (LiCor 6262). We found that during the late wet season (April-June 2001) CO₂ fluxes for the 8 frequently sampled chambers averaged about 3.2 μmol CO₂ m^-2 s^-1. Fluxes decreased slowly from the end of the wet season in June through the end of the dry season (November-December) when CO₂ flux averaged was about 2.2 μmol CO₂ m^-2 s^-1. With the onset of the rain in January 2002, fluxes increased rapidly to approximately 3.0 μmol CO₂ m^-2 s^-1. Both soil temperature and soil moisture exert strong controls on soil-atmosphere CO₂ flux. As we would expect, soil CO₂ flux is positively correlated with both soil temperature and soil moisture content. Because soil temperature and soil moisture are negatively correlated with one another, one effect tends to mask the other.