Three fundamental stages and technologies as main parts of a precision farming project should be considered precisely. These are access to actual multi- dimensional variability detail or variable description on farms, creating a suitable variable-rate technology, and finally providing a decision support system. Some results of a long term practical research conducted by the author in Upon-Tyne Newcastle University of UK for reliable yield monitoring and mapping were utilised to prepare this paper.

The objective of this study is preparing a plan to predict variable rated P requirements spatially in frame-work of this new idea. Using the conventional method of uniform rated fertiliser planning within a field or a hectare unit, we can predict the variable required fertiliser (kg/ha) for each crop spatially based on the previous year’s yield mean (t/ha) and soil analysis within each field. In Britain, the soil analysis results of macronutrient such as P, K are classified into 9 indices (MAFF, 1994). Hence, average soil index and previous year’s yield map of each field with or without straw removal were used to predict VRP fertiliser recommendations for the wheat or barley.

As the first trails in England, some tests, due to electronically yield data logging aided by GPS and mapping in GIS, were done on Nafferton farm of Newcastle University. The author tried to use the information and process these data for the above objective. For instance, further processing of the regular normalised yield data was prepared to predict variable rates, phosphorus off-take maps of East-Hemel field in 1995 and 1996. The study showed although the resultant maps, analytical and statistical considerations could be used to determine site-specific P requirements within each small grid of a field, however, for more precise and confidential conditions of accumulation of yield data collection in several years integrated other points such as soil index and landscape, crop and climate factors also may are required subsequently.