| Abstract: | The main agricultural crops where decision support systems (DSS) can be used via the Internet in Latvia are winter and spring cereals and potato. Two PC‐based models, forming part of a Danish DSS, were tested under the agroecological conditions of Latvia in 1999/2002: PC‐Plant Protection to control diseases in cereals and several modifications of the NegFry model for prediction of potato late blight. The results of 4 years of trials suggest that models that satisfy the needs of one pest may not fit another. The main reasons for failure to adapt PC models are differences in cultivar susceptibility, differences in pathogenicity, simultaneous action of other organisms and spatial placement of crops (forests, rivers and fallow land). For example, it is well known that, with the recent global migration of more aggressive strains and populations of Phytophthora infestans, late blight epidemics have become less predictable and, at the same time, less controllable in potato‐growing areas. For cereals, there is a different spectrum of prevailing pathogens, causing different levels of damage, requiring incorporation into models of thresholds corresponding to local conditions. Data from weekly monitoring of local fields, warnings about the local situation and meteorological information via the Internet are the most important computer‐aided elements for experts in plant protection. |
