FIELD AND LABORATORY EXPERIMENTS TO VALID ATE PREDICTORY MODELS OF THE EXPERT-SYSTEM HERBASYS APPLIED TO PROBLEMS OF CROPS FOLLOWING THE USE OF HERBICIDES

Field and laboratory experiments to valid predictory models of the exp ert system HERBASYS applied to problems of crops folio wing the use of herbicides Field and laboratory studies were conducted with the herbi cides Dicuran (chlorotoluron) and Tramat (ethofumesate) to investigate their breakdown behaviour and effects on subsequent crops. These stud ies served to validate the simulation of breakdown and the predictions for subsequent crops of the expert system HERBASYS. The parameters ne cessary for the simulation of breakdown were ascertained under differe nt incubation conditions for various soils (clay loam, loamy sand, loa my clay) and with the aid of an iterative method of assessment (WALKBA CK). The breakdown of both active substances was accelerated by increa ses in soil temperature and moisture, insofar as the optimum levels we re not exceeded. In the field, breakdown at the various sites was depe ndent on weather conditions but no clear relationship with sorption-de termining properties could be estabilished. The simulated residues acc orded well with those measured in the field and it is possible to tran sfer the breakdown parameters determined for one chemical-soil combina tion to other site and climate conditions. In long-term bioassays with nonsorptive substrates dose-effect relationships were established bet ween chlortoluron and ethofumesate and various crop plants and the rea ction of the crops to the herbicides was investigated in field tests. The effects of available residues on subsequent crops, as calculated f rom sowing date and sorption, were compared with those known from the dose-effect relationships. In general, the risk to subsequent crops wa s well forecasted.