Skip to main content
ARS Home » Research » Publications at this Location » Publication #81630

Title: MULTI-YEAR EVALUATION OF MODEL-BASED WEED CONTROL UNDER VARIABLE CROP AND TILLAGE CONDITIONS

Author
item HOFFMAN, MELINDA - IOWA STATE UNIVERSITY
item Buhler, Douglas - Doug
item OWEN, MICHEAL - IOWA STATE UNIVERSITY

Submitted to: Journal of Crop Production
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/3/1998
Publication Date: N/A
Citation: N/A

Interpretive Summary: Recent public concern over environmental pollution from farming practices has stimulated discussion on methods to reduce herbicide use in crop production. Decision-support models are beginning to provide tools to improve farm profits while holding the promise of reducing herbicide use. Bioeconomic models integrate biological processes of weed-crop interactions with economic factors to give farmers more comprehensive information to guide weed control decisions. While several decision-support models have been developed, field testing has been limited. The goal of our research was to evaluate a bioeconomic weed management model for corn under a range of field conditions. Our results showed that model-generated treatments reduced herbicide use compared with a standard farmer practice in the initial year of the study. However, in succeeding years, weed densities increased in model-generated treatments resulting in greater herbicide use and reduced crop yields compared with standard practices. The results of this research suggest that short-term reductions in herbicide use may have longer-term consequences if weed control levels are not maintained. The bioeconomic model showed potential to control weeds with less herbicide and at a lower cost to farmers. However, it will be important to integrate bioeconomic models in the broader context of weed management to prevent weed population increases that may increase herbicide use or reduce crop yields.

Technical Abstract: Field research was conducted from 1994 to 1996 to evaluate a bioeconomic model used as a weed control decision aid. The model selects pre-emergent weed control tactics based on weed seed bank size and post-emergent strategies based on weed seedling densities. Model-based and standard-herbicide weed control tactics were compared in continuous corn with a long-term history of differing tillage and weed management regimes. Fewer weed seeds were found in plots that were conventionally tilled and no-tillage tended to increase seed bank size. Using standard-herbicide weed control reduced numbers of weed seeds in the seed bank each year compared with model-based weed control. Variation among model-based weed control treatments applied to plots with different cropping was eliminated over the course of the study as economic considerations seemed to predominate over those of control efficacy. As a result, the abundance of foxtail seedlngs relative to that of other species rose to 98% after the third year in model-based weed control treatments compared with 70% in standard herbicide treatments. There were residual effects on weeds due to cropping methods used in the past, but it did not affect corn yields. In 2 of 3 years, corn yields were less with model-based compared with standard weed control. Because weed densities were greatest in the year weed control did not affect yield, we concur with others that the successful use of models as weed control decision aids depends on weed pressure. Data are needed that validate and parameterize models under varying weed levels and cropping practices and take fuller account of weed seed production from sub-threshold levels of weeds.