Monday, November 5, 2007
90-4

Ontology-Based Simulations for Minimizing Environmental Impacts of Florida Sugarcane Production.

Ho-young Kwon, Sabine Grunwald, Howard W. Beck, and Kelly T. Morgan. Soil and Water Science Department, University of Florida, 2169 McCarty Hall, PO Box 110290, Gainesville, FL 32611

A Nutrient Management Plan Support System (NUMAPS) is a decision support system incorporating record keeping and computer simulation that has been developed to assist Florida's citrus growers in selection and application of best management practices for nutrient management. The computer simulation utilizes a data modeling environment where the processes related to citrus growth, soil moisture, and nitrogen (N) uptake are represented as database objects in an ontology-based simulation (OntoSim). Computer code (Java) used to run the simulation, is generated automatically from diagrams, equations, and symbols, in order to provide accurate assessments of water and N utilization and potential losses via leaching and/or runoff. Recent efforts have been made to expand NUMAPS from its original form as a water and N balance model on mineral soils for citrus to include phosphorus (P) on organic soils for sugarcane. The processes for soil organic matter decay, which are based on the mathematical framework of the CENTURY model, were included as well as dynamics of inorganic nutrients (N and P). Sugarcane growth and its nutrient uptake were modeled according a modified version of the DSSAT-CANEGRO model. Also, the current water balance model was improved by i) simulating an impermeable layer in the soil profile to model a perched water table and ii) making deep seepage dependent on vertical saturated hydraulic conductivity and water table height, so that the model hydrology can simulate water table control systems commonly managed on Florida sugarcane fields where sub-irrigation and open ditch drainage are used. After completing these modifications a sensitivity analysis of the model performance to variations in specific model parameters and functions was performed.