Phosphate Fertilizer Recommendation Domains for Maize Using Geographical Information Systems (GIS) in Kisii District

Abstract

A comprehensive GIS database was established for fertilizer recommendation domains in Kisii district by using FURP fertilizer trial results, KSS soils data and .MDBP climatic data. These were manipulated in ESRI's (Personal Computer Environmental Systems Research Institute) ARCINFO and ARCVIEW softwares. The extrapolations were only done for the long rains season (March-August) with three to four years data. GIS technology was used in cluster fertilizer recommendation domains as a geographical area expressed in tenns of variation over space and not limited to the site of the experiment where a certain agronomic or economic fertilizer recommendation was made. The extrapolation over space was found to be more representative for any recommendation, the result being digital maps describing each area in the geographical space. GIS techniques made it possible to model and extrapolate the results using the available data. The extrapolations still need to be verified with more ground data from fertilizer trials. Data gaps in the soil map left some soil mapping units with no recommendations. Elevation was observed to influence yields and it should therefore be included in future extrapolations by clustering digital elevation models with rainfall data in a spatial model at the district scale.