A spatio-temporal model for the simulation of weed population dynamics in terms of use for the development of environmentally friendly weed control strategies

Modern agriculture is exceedingly dependent on the use of pesticides. Without pesticides the obtained yields in agricultural production were not possible. However, the use of pesticides, involves some risk for the natural environment and human health. Therefore, application strategies are investigated to minimize the amount of pesticides, especially herbicides applied - and hence, the potential risk - without significantly yield reduction. Site-specific weed control considering economic thresholds is a promising strategy for reducing herbicide use. In this study, spatial dynamics of different weed populations were modelled by coupling a local population model with a cellular automation model. This approach also allowed for the consideration of the dependence of weed dynamics on habitat characteristics. The model was applied to several weed species, Apera spica-venti, representative for a grass weed, as well as Stellaria media, representative for a broad-leaved weed. Different scenarios of site-specific weed control were simulated over different time periods and their impacts on herbicide reduction and the weed population were caluculated and evaluated.