Use of an expert system for integrating weed control strategies in a weed science laboratory.

Integrated weed management decisions require knowledge of weed taxonomy, weed interference effects, and effectiveness of control. Because they are multifaceted, principles of integrated weed management are difficult to teach in the classroom. A series of laboratory exercises have been developed in which weed taxonomic skills were coupled with the use of a modified version of HERB, a computerized economic threshold-based weed management decision-aid. The laboratory series began with 4 wk of weed taxonomy, followed by a pest management exercise in which two simulated weedy soybean [Glycine max (L.) Merr.) fields (in the greenhouse) were studied. Weed species, density, and size were determined for each field. This information, along with cultivation method, herbicide used, application cost and method, and yield information, serve as inputs to HERB. After inputing information, the decision-aid program then computes the cost of not controlling the weed population (damage estimate). This cost is compared with the costs of available control measures, and appropriate control measures are ranked in order of net-profitability. The student takes the decision-aid output and interprets the results of the damage estimate computations, and its relationship to the costs of control as they relate to the economic threshold principle. In a survey conducted following the lab sequence, 93% of the students agreed the laboratory series increased student appreciation for the importance of weed identification competency. In addition, 87% of the students agreed use of HERB made weed interference and economic threshold principles presented in lecture more meaningful. RECENT economic and environmental concerns coupled with an ever-expanding number of weed control approaches have presented the student, teacher, and crop manager with a perplexing array of information upon which weed control decisions are made. As concerns about production efficiency and environmental impact are heightened, more factors must be considered by practicing weed scientists beD.A. Mortensen, Dep. of Agronomy, Univ. of Nebraska, Lincoln, NE 68583-0915; H.D. Coble, Dep. of Crop Science, North Carolina State Univ., Raleigh, NC 27695-7620; and J.R. Smart and T.A. Bauer, Dep. of Agronomy, Univ. of Nebraska, Lincoln, NE 685830915. Contribution from the Nebraska Agric. Res. Div. Journal Paper no. 8988. Received 31 July 1989. "Corresponding author. Published in J. Agron. Educ. 19:181-183 (1990). fore control decisions are made. It is important that classroom and laboratory instruction in weed science address principles underlying best management practices so students have a clear understanding of how to improve production efficiency, reduce undesirable environmental impacts, and incorporate the most costelfective control strategies. Computers represent a promising tool for integrating large amounts of data to arrive at economically and environmentally sound weed management recommendations. Computer-based management programs have been effective tools for processing large amounts of information. These programs also have been used to enhance student understanding of fertilizer rate response theory (2, 6), turfgrass pest management (4, 5), apple pest management (1), and crop management (7). Such computer programs are effective because they integrate biologic, agronomic, and economic information in a way that is very difficult to do in the classroom (7). The hardware and software are presently available to aid in integrated weed management decision-making. This article will outline how such a program was integrated in a weed science laboratory sequence. The purposes of this article are to: (i) outline the way in which weed taxonomy and an integrated weed management decision-making program were combined in a weed science laboratory exercise, (ii) review the decision-aid program used, and (iii) review student response to the laboratory exercise.