On the decision-making side DSSAT/CropDSS, LUCIA and WaNuLCAS run
on predefined crop rotations, that is, their land use and management options do not
dynamically react to the biophysical model component. FALLOW and the Common Resources Multi-Agent System (CORMAS) build their decision-making rules
on decision trees; the former aggregates at the landscape scale while the latter uses
agents. Mathematical Programming-based Multi-Agent Systems (MP-MAS) uses
mathematical programming to maximize the net income of farm households.
10.1.3 Overview
The contributions in this chapter are roughly ordered by the level of integration they
represent, needed in each case to address a specific research question.
Section 10.2 presents LUCIA which was developed within the Uplands Program
to integrate matter fluxes (lateral subsurface water flows, surface run-off, erosion,
deposition and leaching) between upland and lowland areas in the landscape.
LUCIA is applied here to assess the impact of soil conservation measures on
maize yields in the Chieng Khoi watershed, Vietnam. While LUCIA cannot handle
large watersheds due to the large amount of spatial interactions calculated on high
resolution pixels, Sect. 10.3 expands the scenarios from the plot to the (sub-)
national level and shows that the Crop System Model – Decision Support System
for Agrotechnology Transfer (CSM-DSSAT) and its spatial extension, CropDSS,
can be used to assess the impact of climate change on paddy rice production in
Thailand.
Section 10.4 gives an example of qualitative data obtained from participatory
interviews, decision trees and remote sensing in order to parameterize a semiquantitative land use model. The study highlights the stepwise calibration and
validation of the model building on stakeholder feedback, projecting socioeconomic and biophysical trends over longer time periods and showing, for the
case of Chieng Khoi, how increasing crop production levels masked a steady
decline in inherent soil fertility. What appeared sustainable from an economic
perspective owed to a combination of soil mining plus new varieties and increasing
fertilizer application. Land use sustainability in a smallholder village in north-west
Thailand is assessed in Sect. 10.5, based on predefined sustainability criteria. The
CatchscapeFS approach is a heuristic decision model with an underlying decision
tree structure, and, the approach presented in Sect. 10.4 – CatchscapeFS-CORMAS,
can facilitate participatory approaches such as companion modeling, that is, the
application of a model by farmers, with the guidance of a researcher.
The contribution in Sect. 10.7 leads back to the more academic level. Within the
Uplands Program MP-MAS has been the most widely used instrument for assessing
household (agent) decision-making, but is complex and data demanding and therefore less suitable for participatory approaches. Being a fully fledged economic and
learning model, some applications of MP-MAS use empirical biophysical functions
provided by the TSPC. In a case study of the Mae Sa watershed, Thailand, the
10 Integrated Modeling of Agricultural Systems in Mountainous Areas
371
on predefined crop rotations, that is, their land use and management options do not
dynamically react to the biophysical model component. FALLOW and the Common Resources Multi-Agent System (CORMAS) build their decision-making rules
on decision trees; the former aggregates at the landscape scale while the latter uses
agents. Mathematical Programming-based Multi-Agent Systems (MP-MAS) uses
mathematical programming to maximize the net income of farm households.
10.1.3 Overview
The contributions in this chapter are roughly ordered by the level of integration they
represent, needed in each case to address a specific research question.
Section 10.2 presents LUCIA which was developed within the Uplands Program
to integrate matter fluxes (lateral subsurface water flows, surface run-off, erosion,
deposition and leaching) between upland and lowland areas in the landscape.
LUCIA is applied here to assess the impact of soil conservation measures on
maize yields in the Chieng Khoi watershed, Vietnam. While LUCIA cannot handle
large watersheds due to the large amount of spatial interactions calculated on high
resolution pixels, Sect. 10.3 expands the scenarios from the plot to the (sub-)
national level and shows that the Crop System Model – Decision Support System
for Agrotechnology Transfer (CSM-DSSAT) and its spatial extension, CropDSS,
can be used to assess the impact of climate change on paddy rice production in
Thailand.
Section 10.4 gives an example of qualitative data obtained from participatory
interviews, decision trees and remote sensing in order to parameterize a semiquantitative land use model. The study highlights the stepwise calibration and
validation of the model building on stakeholder feedback, projecting socioeconomic and biophysical trends over longer time periods and showing, for the
case of Chieng Khoi, how increasing crop production levels masked a steady
decline in inherent soil fertility. What appeared sustainable from an economic
perspective owed to a combination of soil mining plus new varieties and increasing
fertilizer application. Land use sustainability in a smallholder village in north-west
Thailand is assessed in Sect. 10.5, based on predefined sustainability criteria. The
CatchscapeFS approach is a heuristic decision model with an underlying decision
tree structure, and, the approach presented in Sect. 10.4 – CatchscapeFS-CORMAS,
can facilitate participatory approaches such as companion modeling, that is, the
application of a model by farmers, with the guidance of a researcher.
The contribution in Sect. 10.7 leads back to the more academic level. Within the
Uplands Program MP-MAS has been the most widely used instrument for assessing
household (agent) decision-making, but is complex and data demanding and therefore less suitable for participatory approaches. Being a fully fledged economic and
learning model, some applications of MP-MAS use empirical biophysical functions
provided by the TSPC. In a case study of the Mae Sa watershed, Thailand, the
10 Integrated Modeling of Agricultural Systems in Mountainous Areas
371