| A weed management support system for Arkansas rice production was developed. The system is a composite of expert systems and compiled BASIC programs. The objective of the system was to assist producers in making weed control decisions by providing, for a given set of field conditions, several weed control recommendations with their associated economic analyses. This would enable the producer to consider both economic and noneconomic factors when making weed control decisions.;In making recommendations, the system considers: the weed complex information, the rice variety, the development stage of the rice, the water management status of the field, and the need for residual herbicide activity.;The system's economic analysis consists of several models, including a model to estimate the rice yield reduction for the 40 different weeds species covered by the system. The inputs for the models include: the species present in the weed complex, the densities and ages of the weeds, the age of the rice, the number of days remaining in the season, the effectiveness of the treatment in controlling the individual weeds, the costs associated with applying the chemicals recommended by the treatment, and the rice harvesting and transportation cost. The information determined by the models include: the impacts of the weed complex on rice yield and rice quality both with and without the proposed treatment; the cost to apply the treatment; the additional harvest and transportation cost; and the economic return for the treatment.;Since the system is intended to be operated by producers on their own personal computers, the user-interface was designed to make the system simple to operate. The interface mechanisms consists of pull-down menus, multiple choice questions, fill-in-the-blank questions, spreadsheets, and text-viewing windows. The system's help features include both context sensitive help, and an operational message line on the screen.;Preliminary testing and evaluation of the expert systems, the economic calculations, and the user interface indicate that the system is functioning correctly and generating reasonable results. Accordingly, the system will be distributed to a limited number of producers for pilot testing in the 1992 growing season. |
