Focus 3

"The Sulawesi Throughfall Displacement Experiment - Ecosystem and economic responses to ENSO droughts in rainforest and agroforest"

Introduction
Most droughts in the non-seasonal rainforest region of Southeastern Asia are related to the El Niño Southern Oscillation (ENSO) phenomenon, which occurs in average every seven to eight years. Several climate scenarios predict more frequent occurrence of ENSO droughts with increasing severity induced by global warming. Detailed assessments of the ecological consequences of droughts in perhumid forests and cacao agroforestry systems are scarce and knowledge whether and how these ecosystems are adapted to severe droughts is limited.

Study area & experimental design
For our study within the DFG funded SFB 552 “STORMA” we selected a perhumid montane primary rainforest and a cacao agroforestry system in Central Sulawesi, Indonesia (Fig.1). The studied rainforest is part of the Lore Lindo Nationalpark. As ENSO events can only hardly be predicted, we simulate drought periods as they appear during middle and strong ENSO events in the study area. With special roof installations we collect about 2/3 of the throughfall water in the forest and cacao sites and channel it with gutters out of the roof plots. We installed 3 roofs in the rainforest (40 x 40 m) and 3 in the cacao agroforest (40 x 35 m), with three control sites next to the roofs in each ecosystem. In each plot 3 soilpits down to 3 m depth were excavated. Around all plots a 40 cm deep trenching was conducted. Weak ENSO
events as we observe this year have only minor effects in our study area. So in February 2007 we will start to simulate a first desiccation period of about 4 months, installing removable plastic-covered bamboo-panels on a wooden gutter construction (Fig. 2 & 3). After a rewetting phase of about 4 months we plan to simulate a more severe ENSO drought event of probably 8 months.

Objectives & Hypotheses
Central objectives are (1) to analyze if and how well agroforestry systems and rainforest are adapted to ENSO droughts, (2) how local and regional economic systems respond to ENSO droughts, and (3) how well ENSO events and their effects can be deduced in the past, measured at present and be predicted for the future in Central Sulawesi.
The main hypotheses are (1) that ENSO droughts affect forest stability in the study area, (2) natural forest trees in the perhumid study area are less well adapted to ENSO droughts than are trees in tropical climates with short or extended seasonal droughts, (3) yields and stability of cacao agroforestry systems are strongly affected by ENSO droughts, (4) the largest economic impact of ENSO droughts is on reduced profits from cacao agroforest, and (5) social and technological innovations to reduce vulnerability to ENSO drought effects are stronger influenced by social and personal capital than by economic necessity.

Methods
With numerous methods we study the drought effects and processes in agroforest and rainforest. We list here the central and main measurements we conduct:
Climate: Net of weather stations, microclimate and throughfall measurement above and below roofs
Soil: Soil physics (Fig. 4a.), water content & potential, water nutrient contents, CO2 concentration (Fig. 4c.)
Roots: Distribution, nutrient content, -activity and vitality, sap flow (Fig. 4b.), water potential, hydraulic conductivity, cavitations frequency
Stems and twigs: Stem diameter increment, xylem sapflow in stems; water potential, hydraulic conductivity and cavitations frequency of twigs
Leaves: Litter production and decomposition, nutrient content, water potential, canopy structure (Fig. 4.d)
Socio-economics: Cacao yield measurement, household surveys, assessment of vulnerability to poverty caused by ENSO droughts, and of social adaptability