Modeling land use change from smallholder agricultural intensification
Agricultural expansion in the rural tropics brings much needed economic and social development in developing countries. On the other hand, agricultural development can result in the clearing of biologically-diverse and carbon-rich forests. To achieve both development and conservation objectives, many government policies and initiatives support agricultural intensification, especially in smallholdings, as a way to increase crop production without expanding farmlands. However, little is understood regarding how different smallholders might respond to such investments for yield intensification. It is also unclear what factors might influence a smallholder’s land-use decision making process. In this proposed research, I will use a bottom-up approach to evaluate whether investments in yield intensification for smallholder farmers would really translate to sustainable land use in Indonesia. I will do so by combining socioeconomic and GIS data in an agent-based model (Land-Use Dynamic Simulator multi-agent simulation model). The outputs of my research will provide decision makers with new and contextualized information to assist them in designing agricultural policies to suit varying socioeconomic, geographic and environmental contexts.
Eric Kameni holds a Ph.D. in Computer Science option modeling and application from the Radboud University of Nijmegen in the Netherlands, after a Bachelor’s Degree in Computer Science in Application Development and a Diploma in Master’s degree with Thesis in Computer Science on “modeling the diffusion of trust in social networks” at the University of Yaoundé I in Cameroon. My doctoral thesis focused on developing a model-based development approach for designing ICT-based solutions to solve environmental problems (Natural Model based Design in Context (NMDC)).
The particular focus of the research is the development of a spatial and Agent-Based Model to capture the motivations underlying the decision making of the various actors towards the investments in the quality of land and institutions, or other aspects of land use change. Inductive models (GIS and statistical based) can extrapolate existing land use patterns in time but cannot include actors decisions, learning and responses to new phenomena, e.g. new crops or soil conservation techniques. Therefore, more deductive (‘theory-driven’) approaches need to be used to complement the inductive (‘data-driven’) methods for a full grip on transition processes. Agent-Based Modeling is suitable for this work, in view of the number and types of actors (farmer, sedentary and transhumant herders, gender, ethnicity, wealth, local and supra-local) involved in land use and management. NetLogo framework could be use to facilitate modeling because it portray some desirable characteristics (agent based and spatially explicit). The model develop should provide social and anthropological insights in how farmers work and learn.