Agricultural models have been commonly used to extend the results of experimental research to other soil types, climates, and management conditions outside the experimental design. For example, they have been used for extrapolating limited duration experimental results to variability in climatic conditions across longer periods of time (e.g., 25-50 years), and to extreme climatic conditions (e.g., droughts or flooding) not encountered during the study period. Agricultural models have proven to be useful tools for in-depth analysis of problems in management, environmental quality, global climate change, and other ecological issues, and can thus be a basis for policy or regulatory use. Agricultural models also function as decision aids in choosing best management practices for long-term sustainable production, as well as helping to guide site-specific management on agricultural landscapes and within-season dynamic management in response to spatially variable soil moisture and weather conditions. Many natural resource DSSs have an agricultural model at their core, but are also supported by soil, climate, and management databases, environmental and economic analysis packages, user-friendly interfaces to check default data or enter site-specific data, and graphical visualization of simulation results. An example is the design of the USDA-ARS GPFARM DSS (Figure 2). GPFARM is a whole-farm DSS for strategic planning and evaluation of cropping systems, range-livestock systems, and integrated crop-livestock farming options for production, economics, and environmental impacts.
Complex and highly detailed process-level agricultural models are generally too difficult for consultants, producers, or policymakers to directly use. An alternative approach is to create an integrated research-information database as a DSS core in place of a simulation model. An agricultural system model, evaluated against available experimental data, is used to generate production and environmental impacts of different management practices for soil types, weather conditions, and cropping systems outside the experimental limits. The model-generated
(Graphical user interface)
Crops (animals) Fertility Pests
(Soil-crop-)animal simulation model
Simulate on-farm production Estimate environmental effects Calculate required inputs for other modules
Land management unit
Soil and land-use database
Calculate gross margin of return
Determine net farm income Estimate economic risk of current enterprises
Annual netfarm income
Crops and animals
GPFARM information system (http://infosys.ars.usda.gov)
Support and maintenance
Enterprise database of le ity ide nt tici ua st ue o a.
Estimate off site effects of the chemicals being used Indicate current on-farm production capability and future sustainability
Mobility, toxicity, persistence
Water quality index
Parameters describing acceptable environmental impacts
Output visualization (model output responses, indices, MCDM analysis)
information is then combined with experimental data and long-term management experience of farmers and field professionals to create a database. These databases are often then combined with a socioeconomic analysis package or other tools (e.g., multiobjective decision analysis) in order to conduct a tradeoff analysis between conflicting objectives such as economic return and environmental quality. Overall, this type of approach is very flexible in generating site-specific management recommendations and avoids the problem of having to interpret complicated model output.
Was this article helpful?