PRESCRIPTION MAPS FOR SPATIALLY-VARIABLE HERBICIDE APPLICATION IN NO-TILL CORN

Weed maps for fields of no-till corn (Zea mays L.) were prepared from image analysis of digitized low-altitude aerial photographs. These wee d maps were imported into a Geographic Information System (GIS) and di vided into independent subunits for spatially variable herbicide presc ription. A decision model was designed for pre-plant and post-emergenc e weed control recommendations. Each subunit of the field weed map was submitted to this decision model to determine the optimum herbicide m ix and application rate. The resulting prescription maps would be used to control a field sprayer and to apply the appropriate herbicide com bination to each weedy area. Results indicate that herbicide use would have been reduced by more than 40% with this approach. This demonstra tes a means to significantly reduce herbicide usage in crop production without sacrificing weed control or crop yield.