Decision support for variable liming rates

Abstract

Land areas may have considerable variation in characteristics that affect crop production such as different soil types or fertility levels. For optimum farm management this spatial variability should be taken into account. The thrust of this project was to examine information and information management systems for use in spatially variable farmland management. The variability of soil acidity and the addition of limestone to counteract that acidity were the components studied. Intensive soil sampling of a test field in East Texas showed substantial spatial variability of soil acidity, with pH having a mean of 5.5 and standard deviation of 0.42. The data were analyzed using geostatistical methods to determine experimental semivariograms and to krige best fit maps for levels of pH, soil texture and buffer pH. Map data were then combined using geographic information system (GIS) techniques to develop variable rate liming application plans. Analysis showed that using the mean liming recommendation was less than local liming requirements (would cause an under-application) on approximately 37-41% of the field, depending upon the liming recommendation method used. Geographic Information Systems (GIS) are computerized systems designed to manage mapped data. A key to the adoption of variable rate farming techniques will be the ability to examine and analyze information. A survey of microcomputer based GIS with the potential for on-farm use was conducted. Nineteen GIS systems were summarized. Computer simulation was used to determine the constants of a control system compensator. The output response of a granular fertilizer spreader control system, such that would be used for a lime applicator, was evaluated using computer simulation.