Modeling spatial and temporal dynamics of Chenopodium album L. under the influence of site-specific weed control
- Publication Type
- Journal contribution
- Authors
- D. Dicke and R. Gerhards and A. Büchse and K. Hurle
- Year of publication
- 2007
- Published in
- Crop Protection
- Band/Volume
- 26/3
- ISBN / ISSN / eISSN
- 0261-2194
- Page (from - to)
- 206-211
Experiments were conducted on four arable fields in order to study the spatial and temporal dynamics of weed populations in a 4-year crop rotation under the influence of site-specific weed control from 1997 until 2003. Winter wheat (ww), winter barley (wb), maize and sugar beet (sb) were rotated in the experimental fields. The objective of the study was to adapt a modeling tool based on Zwerger and Hurle [1990. Untersuchungen zur Abbildungsgüte simulierter Befallsverläufe bei Unkräutern. (Experiments for closeness of simulated weed infestation runs). J. Plant Diseases Prot. 97, 133-141] to forecast weed population dynamics taking into account site-specific weed control. For weed mapping, a regular 157.5 m grid was established in all fields. Weed seedling density was counted before and after post-emergent herbicide application and before harvest in a 0.4 m2 quadrate frame placed at all grid intersection points. Application maps were created to direct a GPS-controlled patch sprayer. The patch sprayer varied the herbicide dosage and mixture depending on the information in the application maps. Seed production of surviving weeds and seed mortality by predation and fatal germination was assessed in field studies. Data that were needed for the model and not assessed in the experiments were taken from literature. All data, including site-specific weed control, were entered in the weed population model. Distribution of Chenopodium album L. in two fields was predicted based on previous years weed distribution maps and compared to the observed data. In both fields weed density of C. album did not increase during the 6 years of site-specific weed control. Predicted and observed density values were significantly correlated using Pearson's correlation coefficient. Modeling improves our understanding of spatial and temporal dynamics of weed population and thus can be included in decision algorithms for patch spraying.