Eliciting farming decisions of smallholders in response to water availability conditions in Chingale, southern Malawi. In R. Seppelt; A. A. Voinov; S. Lange; D. Bankamp (Eds.). Managing resources of a limited planet: proceedings of the International Congress on Environmental Modelling and Software (iEMSs), Sixth Biennial Meeting, Leipzig, Germany, 1 - 5 July 2012.

The objective of this project is to examine adaptation strategies of smallholders in the Chingale district of Malawi to climate change impact, as well as to the implications of introducing the Integration of Aquaculture into Diversified Food production Systems (IADFS). The central approach of this research is combining an agent-based model (ABM) representing farmers with various types of production systems with a water resources model (WRM). Whereas the WRM relies on historical time series data of rainfall/runoff and a set of climate change scenarios, the ABM builds upon recent household and village size surveys in combination with population census and GIS data. However, knowledge that feeds into the decision rules for the model can only come from the farmers in an interactive way. For this reason a knowledge elicitation workshop was conducted where farmers simulated their main farming activities on a monthly basis. As the primary driver for decision making, water availability was employed. Water availability was simulated to change on a monthly base during both a year with adequate water availability and a year with insufficient water availability. The most relevant production systems were examined during four workshop sessions: rain-fed farms (RF), canal irrigation farms (IR), farmers in the low-lying areas with high ground water tables and adapted cultivation methods (Dimba) and integrated agriculture aquaculture farms (IAA). This simulation workshop resulted in a variety of insights such as: a diversity of individual (drought) coping strategies such as postponing sowing activities, reducing land devoted to crops, or reducing irrigation. Furthermore, the assumption that rain-fed farms are most sensitive to droughts was confirmed, whereas groundwater-fed farmers demonstrate the lowest sensitivity. The resulting decision rules will be validated with farmers and experts of the regional farm systems in a later phase of the project