Evaluation of economic outcome associated with a given weed management system is an important component in the decision-making process within crop production systems. The objective of this research was to investigate how risk-efficiency criteria could be used to improve herbicide-based weed management decision making, assuming different risk preferences among growers. Data were obtained from existing weed management trials in corn conducted at the University of Minnesota Southern Research and Outreach Center at Waseca. Weed control treatments represented a range of practices including one-pass soil-applied, one-pass postemergence, and sequential combinations of soil and postemergence herbicide application systems. Analysis of risk efficiency across 23 herbicide-based weed control treatments was determined with the mean variance and stochastic dominance techniques. We show how these techniques can result in different outcomes for the decision maker, depending on risk attitudes. For example, mean variance and stochastic dominance techniques are used to evaluate risk associated with one- vs. two-pass herbicide treatments with and without cultivation. Based on these analyses, it appears that a one-pass system is preferred by a risk-averse grower. However, we argue that this may not be the best option considering potential changes in weed emergence patterns, application timing concerns, etc. The techniques for economic analysis of weed control data outlined in this article will help growers match herbicide-based weed management systems to their own production philosophies based on economic risk.

Nomenclature: Corn, Zea mays L.

Additional index words: Analysis of variance, mean variance procedure, stochastic dominance.

Abbreviations: CV, coefficient of variation; FSD, first-degree stochastic dominance; POST, postemergence; PRE, preemergence; SSD, second-degree stochastic dominance.