B6 - Carbon dioxide production in soils under cacao agroforestry and natural forests: effects of drought and landscape variability

Abstract
The main goals of this sub-project are:

• To estimate the effects of natural and experimental droughts and consecutive rewetting on the soil CO2 production and efflux in natural forest and cacao agroforestry ecosystems.
• To identify the environmental controls on spatial variation in soil CO2 efflux in the natural forest during the wet and the dry season.

Central hypotheses for the first goal of B6 are: a) the natural forest has mechanisms like deep roots to reduce impacts of ENSO droughts. b) The cacao agroforestry system is not well adapted to ENSO droughts. The hypothesis to test for the second goal is that a considerable part of the naturally occurring spatial variation in soil CO2 efflux can be explained by landscape and soil properties. We will use a combination of measurements of CO2 efflux and concentrations, with modelling of CO2 transport in soils and statistical models to test these hypotheses. Our Indonesian counterpart plans to study the effects of an experimental drought on CH4 and N2O dynamics.

Summary
Global warming may increase the frequency and intensity of El Niño Southern Oscillation (ENSO) events which cause severe droughts in Indonesia. This effect may be enhanced by the reduced recirculation of water between the biosphere and the atmosphere as a result of deforestation. Changes in rainfall regime will have a direct effect on root dynamics and soil organic matter decomposition, the two processes that contribute to production of CO2 in the soil. Soil CO2 efflux (or soil respiration) constitutes one of the largest CO2 fluxes from the terrestrial biosphere to the atmosphere and is a major component of the global carbon cycle. Although soil CO2 efflux has been the subject of many studies in recent years, it remains difficult to deduct controls of this flux because of the different sources that produce CO2 and because potential environmental controls like soil temperature and soil moisture often co-vary. In addition, spatial variability of soil CO2 efflux, caused by variations in landscape, soils and vegetation introduces a considerable level of uncertainty in modelling soil respiration at landscape and larger regional scales. The main goals of this sub-project are:

- To estimate the effects of natural and experimental droughts and consecutive rewetting on the soil CO2 production and efflux in natural forest and cacao agroforestry ecosystems.
- To identify the environmental controls on spatial variation in soil CO2 efflux in the natural forest during the wet and the dry season.

Central hypotheses for the first goal of B6 are: A) the natural forest has mechanisms like deep roots to reduce impacts of ENSO droughts. However the forest in this perhumid area is not as well adapted as tropical forests in areas with a strong dry season. B) the cacao agroforestry system is not well adapted to ENSO droughts. An experimentally imposed drought could cause an increase in allocation of C to deep roots. If changes in C allocation are substantial, I expect to see greater production of CO2 at depth under the drought stressed treatment. Furthermore, I expect that this effect will be stronger in the natural forest than in the cacao. We will do surface CO2 efflux measurements (using an infrared gas analyzer) together with measurements of CO2 concentration (using a simple GC with TCD detector) throughout the soil profile. An existing model of soil gas diffusivity will be calibrated (using naturally occurring 222Rn concentrations), in which CO2 production will be calculated as a function of soil depth based on soil porosity, water-holding properties, and water content. The hypothesis for the second goal is that a considerable part of the naturally occurring spatial variation in soil CO2 efflux can be explained by landscape and soil properties. We will do a systematic spatial sampling of CO2 efflux at the footprint of the tower in Bariri and an additional natural forest. At each site, measured CO2 fluxes at 100 points across variations in soil and landscape will be used to develop a statistical model to determine the spatial controls on CO2 efflux. An additional set of 100 point measurements will be used to test the quality of the developed model. Separate models will be developed for the wet and dry seasons of the two forest areas. Models will be compared to deduct general spatial and seasonal controls on soil CO2 efflux.