Modelling Herbicide Use in Genetically
Modified Herbicide Resistant Crops - 2 SummaryThe objective of this study is to substitute perennial experiments by simulating the consequences of herbicide use in a crop rotation with transgenic herbicide tolerant sugarbeet or winter oilseed rape in comparison to traditionally grown varieties of these crops. We present two simple empirical models which aim to address in particularly the accumulated use of herbicides and selection of herbicide resistant weeds and volunteers. The first computer simulation model investigates the potential impact of herbicide resistant sugarbeets on future herbicide use in Denmark. The model uses a sigmoid function to simulate daily changes in dry matter of crops and weeds which compete for maximum dry matter production determined by the site potential. Decisions to control weeds are based upon critical levels of dry matter. The dynamics of the weed populations is determined by the efficacy of weed control and changes in the seed bank. Simulations of a rotation with sugarbeet-barley-wheat-wheat showed that herbicide use can be decreased through the use of glufosinate or glyphosate resistant sugarbeet. The second simulation model investigates the potential impact of herbicide tolerant winter oilseed rape on future herbicide use. The model also uses a sigmoid function to simulate daily growth of crops and weeds which compete for a maximum yield potential. Thresholds for weed control are based upon critical levels of weed biomass. The dynamics of the weed population are determined by the efficacy of the herbicide on the individual weed species and seed bank parameters. Herbicide efficacy at different doses is determined by a log-logistic dose-response curve for each species. Simulation of a rotation with winter varieties of oilseed rape-wheat-wheat-barley showed ambiguous results for the two Danish defined measures of herbicide use, because herbicide use in the rotation with glyphosate or glufosinate tolerant oilseed rape was not reduced in kg a.i. ha-1 compared to a traditional treatment whereas treatment frequency (number of standard recommended rates) decreased. The models described in this report are first attempts to integrate knowledge about herbicide tolerant crops with known agricultural practices. It should be emphasized that the models are preliminary and need validation before any reliable predictions can be made about long-term consequences. The models are, furthermore, restricted to simulate one crop rotation with sugarbeet and one rotation with oilseed rape. The predictions from the models are therefore preliminary and only valid for the chosen combination of crops, weeds and herbicide strategies. |