System modelling in plant disease management : barley leaf rust as a case study

The role of system modelling for disease management has been investigated. Static mathematical models have been used to characterise components of the barley leaf rust epidemic as well as the complete epidemic. The development of a dynamic simulation model of the leaf rust-barley system from experimental data is described. This dynamic model accurately predicted disease progress and the yield reduction due to disease. An initial system model which represented a general leaf rust epidemic was constructed and used as a guide for the experimental phase of the investigation. Model results showed that latent period, spore production rate and penetration ratio were critical factors affecting epidemic development. Experiments to elucidate the epidemiology of barley leaf rust, caused by Puccinia hordei Otth, showed that: -the field epidemic was a polycycle resulting from an integration of several monocycles over time. -field epidemics at Lincoln reached higher terminal severities with crops sown later in the season. The mean daily temperature and the number of dew days during an epidemic were found to be positively related to epidemic severity. -increasing ambient temperature in the range 10°C to 20°C increased total urediniospore production per infectious period per urediniurn, decreased latent period and had no effect on infection ratio. Increasing urediniurn density in the range 1.0/cm² to 160.5/cm² decreased the total urediniospore production per infectious period per uredinium decreased latent period and increased infection ratio. -the number of urediniospores liberated increased with wind speed in the range 2.2 km/h to 34.1 km/h. Diurnal periodicity of liberation occurred from changes in environmental conditions. -urediniospores lost their viability with increasing exposure to solar radiation. Under laboratory conditions, urediniospores survived for 38 days. -urediniospores required free moisture to germinate. Increasing the temperature of germination from 9°C to 25°C generally increased the percentage of germinated spores while increasing light intensity from 0 1x to 20500 1x generally decreased germination percentage. -three types of disease-loss models were compared for estimating yield loss caused by barley leaf rust. Of the critical point models, the model I based on disease assessment when the ear has fully emerged accounted for the most variation in yield loss, 79.6%. The best area-under-curve model explained 74.7% of yield loss variation. However, a multiple point model based on assessments at four growth stages was found to be the best predictor of yield loss, explaining 97.6% of the variation in yield loss. The detailed system model of the leaf rust-barley system, BARSIM-I, was successfully validated using data collected from fifteen epidemics in five fields during the cropping seasons of 1975/76 and 1976/77. Three validation procedures were adopted, and each was satisfactory. The management of barley leaf rust was conceptualised as comprising two main activities - projection of disease increase and estimation of yield loss. The integration of these components into a management system was examined in relation to the potential role of BARSIM-I.