|Jallas, Eric - MISSISSIPPI STATE UNIV|
|Whisler, Frank - MISSISSIPPI STATE UNIV|
Submitted to: Journal of Cotton Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: June 13, 2000
Publication Date: N/A
Interpretive Summary: Precision agriculture may increase yields and profits while optimizing inputs and minimizing advance environmental effects. Advances in computers, plant growth models, global positioning systems, and powerful geographic information programs should be linked in a systems approach to precision agriculture. This research linked the ARS cotton model, GOSSYM, the expert system COMAX, and a Geographic Information System from ARCView and applied this system to 88 one ha plots on a grower's farm in the Mississippi Delta. Actual 1997 weather and the grower's production inputs were used to simulate the mean yield of the field based on the predominant soil type, Commerce silt loam. The predicted and actual yields were 996 and 995 lbs/acre. When using actual weather and soil parameters specific to each plot simulations averaged 991 lbs per acre. These simulations showed the effects of variable soil parameters on predicted yields. The linked system was used with 1997 actual weather to optimize nitrogen and water for soil parameters in each plot. These simulations predicted an increase of 289 lbs lint/acre using 27 lbs less nitrogen and 1 inch more water per acre than used by the grower. However, the frequency and timing of water and nitrogen were different than applied by the grower. Managing nitrogen and water for specific soil parameters in this field should be more profitable than using the average soil type for managing the entire field. The research demonstrates how this systems approach can examine variables, select the most promising, and store and display data in a graphic and numerical form. Linking the system with geospatial referenced soil parameters provides information for prescriptions to drive variable rate farm implements that apply variable rates of production inputs.
Technical Abstract: Research in precision agriculture attempts to use a cadre of technologies to make site specific determinations of the amount and timing of seed, agricultural chemicals, irrigation water, and other inputs; thus providing better management of the crop in a manner that should increase profits and also provide better protection for the environment. This manuscript reports how a geographic information system, a plant growth model, and an expert management system were linked and used in a computer based systems approach to examine ways to use precision agriculture in cotton, Gossypium hirsutm L. The ARS cotton model GOSSYM and the expert management system COMAX were linked with a Geographic Information Systems (ARCView) and applied in a systems approach to examine the relative contribution of soil type, water, and nitrogen parameters in precision agriculture for cotton. We compared actual yields and actual weather with system simulations to examine the effects of site specific management of nitrogen and water in a 88 ha field with variable soil types in the Mississippi River Delta of Mississippi. The results indicate that cotton yields in this field should be increased and costs of chemicals should be reduced when nitrogen and water inputs are applied on a site specific basis rather than managing the whole field in a uniform manner. Soil types emerged as the major variable that could be managed more effectively by using rates and timing of nitrogen and water linked to specific soil parameters.