The development of weed management systems requires accurate prediction of weed-crop competition. In this paper, simple regression models of crop yield losses based on weed density and weed leaf area are compared. In weed leaf area models, variations in the relative damage coefficient (q) were also analyzed. Finally, three simple methods to assess weed cover were compared: visual, photographic, and optic device assessment. Leaf area models were at least as accurate as weed density models. However, the generality of the leaf area models was restricted by changes in q, according to the date of leaf area evaluation and the year. Although all methods to assess weed cover correlated adequately with weed leaf area, visual estimates were the best to predict crop yield losses perhaps because very low levels of weed leaf area could be distinguished visually better than by other methods.

Field experiments were conducted in Oklahoma to evaluate the effect of three imidazolinone herbicides and metsulfuron on established field bindweed and hard red winter wheat followcrops. Imazapyr at 280 g ai ha-1 and imazethapyr at 560 g ai ha-1 controlled field bindweed from 78 to 100% for 48 wk, but imazaquin at 560 g ai ha-1, metsulfuron at 17.5 g ai ha-1, and 2,4-D plus picloram at 1120 plus 280 g ae ha-1 did not. Imidazolinone herbicides reduced forage and grain yield of wheat seeded 8 to 14 wk after herbicide application. Only imazapyr reduced grain yield of wheat seeded 15 mo after treatment.

The economic threshold is a concept strongly embedded within the weed management literature. There are some theoretical concerns with applying a static approach such as the economic threshold to weed management decision making. An improvement is to adopt a population management approach where the intertemporal effects of decisions are taken into account. The focus should be on managing weed populations through time rather than minimizing the yield effect of weeds in a single season or year. Rather than viewing weeds as an annual production problem, the weed seed bank can be considered a renewable resource stock, and the management goal is to deplete this resource stock through time. The principles of natural resource economics illustrate that including the intertemporal effects of weed control will, for a given size of a seed bank, result in a greater level of weed control and a higher economic benefit than if control decisions were based solely on the current period effects. A dynamic economic model was developed of an extensive Australian spring wheat (Triticum aestivum) cropping system to test these principles using wild oat (Avena fatua and A. ludoviciana) as an example. The model was solved for a 20-yr time horizon for a population management approach and the traditional static economic threshold. The economic benefits from a population management approach were significantly greater than those generated by the economic threshold, and the final seed bank was considerably lower. This result suggests that a paradigm shift from thresholds to longer term population management is warranted.