A software coupling approach to assess low-cost soil conservation strategies for highland agriculture in Vietnam

Publikations-Art

Zeitschriftenbeitrag (peer-reviewed)

Autoren

Marohn, C., Schreinemachers, P., Quang, D. V., Berger, T., Siripalangkanont, P., Nguyen, T. T., Cadisch, G.

Erscheinungsjahr

2013

Veröffentlicht in

Environmental Modelling and Software

Band/Volume

45/

DOI

10.1016/j.envsoft.2012.03.020

Seite (von - bis)

116-128

Schlagworte

sustainability

Soil degradation is an environmental process mainly caused by land use decision-makers that has substantial feedback effects on livelihoods and the environment. To capture these feedback effects and the resulting human-environment interactions, we used an agent-based modeling approach to couple two software packages that simulate soil, water and plant dynamics (LUCIA), and farm decision-making (MP-MAS). We show that such a software coupling approach has advantages over hard-coded model integration as applied by most other comparable studies, as it facilitates combining of increasingly sophisticated individual models and can achieve a well-balanced representation of agricultural systems. Using a numerical application for a small mountainous watershed in northwest Vietnam we show the challenges in model coupling, calibration and partial validation, and explore the properties of the coupled model system. Scenario analysis covering the introduction of low-cost soil conservation techniques indicates that some of these techniques would have an impact on soil erosion, maize productivity and household income levels in the study catchment area under current conditions. However, maize yields and the adoption of soil conservation appear to be sensitive to the price of mineral fertilizers, with lower fertilizer prices impeding the adoption of soil conservation measures. The software coupling approach was able to capture interactions between decision-makers and natural resources, as well as the level of spatial variability, in more detail than the individual models. Still, the greater number of endogenous variables and thus degrees of freedom increased the importance of validation and testing parameter sensitivity of the results