Prediction of Pests,Diseases and Losses: Principles and Computer Applications1

Efficacy of plant protection is closely related to an appreciation of the seasonal development and population changes of harmful organisms. This necessitates national surveys on pest population dynamics and on the epiphytotics of plant diseases. From these surveys, regional forecasts can be produced to give advice to the farmers. This work needs highly sophisticated systems to cover both numerical and spatial aspects of pest situations. Thus, there is an urgent need to exploit computer applications in solving forecasting problems. However, the use of computers for forecasting the incidence of harmful organisms and losses is in its initial phase. At present, single pest and disease models prevail. Numerical aspects have been approached both empirically and experimentally. Empirical models, based on multiple regression analysis, are relatively simple and easy to use for various harmful organisms. However, simulation models of harmful organisms are expected to lead to much more reliable results, although, so far, little experience is available on their predictive powers in crop protection. Spatial aspects have been approached with a routine computer mapping technique, which seems to be an effective tool for recognizing and forecasting distribution trends, while loss predictions can be based on crop models, incorporating submodels forecasting the incidence of harmful organisms. The prospect for operational use, in forecasting, of the computer techniques mentioned is discussed.