Relationships between crop yields and soil mapping units using a GIS for potential applications in precision farming

The goal of precision agriculture is to divide crop fields into management units of similar production potentials. Without the high resolution crop yield data from yield monitors it has been difficult to document the actual yield of a crop in different parts of a field or include soil mapping units in the development of site specific farming criteria. The objectives of this study were to (1) determine the optimum resolution of yield monitor data for the development of crop yield based management units and (2) evaluate the feasibility of using soil mapping units to develop crop yield management maps. Crop yield data were aggregated and classified within a Geographical Information System for the development of crop yield management units for one field each of corn (131 acres), soybean (45 acres) and wheat (63 acres) within Kent county, Maryland. Methodologies were developed to spatially average crop yields within grids of 1.50, 1.00, 0.50 and 0.25 acres and comparisons were made to determine the effectiveness of this aggregation. Crop yield means of aggregated units were almost identical to means obtained using the original yields. Aggregation into 1.00 acre grids and classification into 20 bushels/acre for corn, 10 bushels/acre for wheat and 5 bushels/acre for soybean represented 8–10% of the average yields and resulted in the crop yield management maps that seemed optimum for field scale management.; The feasibility of using soil mapping units for crop production management was evaluated by testing for differences with original and aggregated crop yields within land areas enclosed by mapping unit delineations using statistical t-tests. The results indicated that though significant differences existed between land areas enclosed by soil mapping unit polygons, there were considerable variation within mapping unit delineations. The absence of adequate information from soil survey data, and the inherent variability associated with crop yields and soil properties made it difficult to directly support the use of soil mapping units in developing crop yield management maps. Field mapping of crop yield variations seemed to have more practical utility over the use of 1:15,840 scale soil survey information for the implementation of spatially selective field operations.