A complex set of interactions among crops, weeds and their environment determines the impact of weed interference on crop productivity. These interactions can be simulated with dynamic crop:weed competition models, such as ALMANAC. In this study, ALMANAC was modi®ed to simulate maize:Abutilon theophrasti competition. In the modi®ed ALMANAC model, daily increases in leaf area index (LAI), height and rooting depth are attenuated on the basis of accumulated above-ground biomass and by environmental stress. Also, a simple,¯exible method is adopted to partition radiation in a mixed canopy. A maize:A. theophrasti competition study conducted near Aurora, NY, in which a range of weed densities (0±16 plants m A2 ) were established in a maize crop, was used to evaluate the model. The modi®ed ALMANAC proved to be a useful tool for segregating the maize response to competition in 1991 (simulated loss of 35% at the highest weed density) from those in 1992±94 (simulated losses not greater than 16%). Based on these ®ndings, the modi®ed ALMANAC model is judged to be capable of distinguishing between environmental conditions that facilitate large yield losses and those that allow maize to outcompete A. theophrasti.

The practical application of single‐season economic thresholds for post‐emergence weed control decisions has been frustrated by the impact of growing season, climate and edaphic factors on crop:weed interactions. Competition in a mixed plant community is a dynamic process that is contingent upon species‐specific growth characteristics and their expression based on environmental conditions. To address these issues, a modified version of the ALMANAC competition model was parameterized and run to simulate maize:Abutilon theophrasti competition with 30 years (1966–95) of historical climate data for a site in New York State. Simulations indicate that, when weeds do not emerge before maize, maize will only suffer substantial yield reductions from A. theophrasti competition in two out of every 10 years at this site. It is also evident that economic thresholds based solely on the level of weed infestation are inherently flawed. Long‐term simulation results suggest that estimates of early season water stress could be used as an independent tool for rationalizing post‐emergence control decisions. Shifting the focus from quantifying the infestation intensity of the weed population to assessing the competitive status of the maize crop indirectly with climate information may alleviate many of the problems commonly associated with threshold management strategies.

No abstract