Abstract
A practical and objective system is needed to determine the lowest rates of the most efficacious herbicides to meet each producer's specific weed control problems. Determining which method of weed control to utilize is difficult today with increasing product choices, the growing use and complexity of herbicide mixtures, regulatory pressures to reduce rates, and the closer integration of weed control with other crop decisions. Expert computer systems could improve current practices and use herbicide mixtures as a tool to increase herbicide efficiency. Such systems would account for herbicide dose and mixture responses; select most economical herbicides; optimize adjuvants; recommend control at economic thresholds; and vary rates according to weed spectrum, density, and local environmental conditions. An example using chlorimuron and thifensulfuron illustrates how these systems could use quantitative dose response and mixture information.