Submitted to: North Central Weed Science Society US Proceedings
Publication Type: Proceedings
Publication Acceptance Date: December 12, 1996
Publication Date: N/A
Technical Abstract: Several bioeconomic weed management computer models have been developed in recent years. However field evaluation of these models has been limited. Research was conducted at Morris and Rosemount, MN from 1991 to 1994 to evaluate a bioeconomic weed management model under field conditions. At Morris, use of the model reduced herbicide application by more than 2 kg/ha/year compared with standard herbicide treatments in corn and soybean. Treatment costs were reduced by $44 to $53/ha and gross margins increased by $54 to $67/ha. Results were more variable at Rosemount. In corn, net return to weed control was not increased by using the model. However, herbicide use was reduced 27% when using the seed bank and 68% when using the seedling model. Frequency of herbicide application was not affected by using the model. In soybeans, use of the model increased net return to weed control compared with a standard herbicide treatment 50% of the time. Herbicide use compared with a standard treatment was reduced by 47% using the seed bank model and 93% using the seedling model. Frequency of herbicide application was not changed by using the model to generate weed control recommendations Overall conclusions from the Morris and Rosemount experiments were that the basic structure of the model is sound. The model was responsive to differing weed populations and usually reduced the quantity of herbicide active ingredient applied. Performance of the model was influenced by weed populations characteristics. Impacts on economic returns were variable among years and locations. Future research with bioeconomic weed management models should include testing under a wider range of weed population and management conditions.