Epidemiology: A logical science to manage plant diseases

Plant disease epidemiology deals with diseases in plant populations. Over the years, it has become a vibrant field of science, achieving significant conceptual innovations with an important impact on the management of plant diseases. Plant disease epidemiologists have a strong scientific tradition in studying climate-pathogen-disease relationships and also in dealing with biodiversity. Understanding the factors that trigger the development of plant disease epidemics is helpful in creating and implementing effective strategies for disease management. Therefore, description and quantification of the disease cycle is the foundation of plant disease epidemiology and the key to developing effective disease management. Disease forecasts which assists growers in determining when to apply disease management techniques, is a logical and exciting application of plant disease epidemiology. It aids growers in making the best decisions among disease management alternatives. Since the development of plant disease epidemics depends upon favourable meteorological conditions, it seems reasonable to examine those conditions associated with a variation in disease severity from year to year. One goal in the analysis of meteorological and disease data is to develop simple models that use a minimum of variables so that the models are easy to use for the management of disease. Also, the prediction models must be based on the meteorological variable that affects disease early enough in the growing season to allow an option of control. Now days, input data for the forecasts are usually derived from automatic weather stations (AWS) located in different regions. In India, late blight forecasting systems for the appearance of the disease have been developed both for hills and the plains and development of the disease support system (DSS) is underway. A computerized forecasting model ‘ JHULSACAST’ has been developed for western UP which has 3 important components- late blight prediction model, a computer programme for forecasting and an interface to utilize weather data from an automatic weather station. The use of computer based disease simulation models is quite common for the predition of important diseases in USA and other developed countries. Downy mildew caused by Pseudoperonospora cubensis is a limiting factor for successful cucurbits cultivation in Punjab conditions and the weather variables like maximum temperature and mean relative humidity during the period corresponding to SMW's 21 to 24 were identified as important components in developing disease prediction model. An early appearance of disease preceding this period helped in build up of initial inoculum and rapid disease take-off if the subsequent conditions were favourable. A function of these two parameters termed as ‘Humid Thermal Ratio’ (HTR) was used to avoid the problem of multicolinearity in regression analysis. Overall the model based on HTR fit the data well as judged by its performance in the validation tests. The forecasts are formulated to be as objective as possible with the meteorological tools available. A degree of human experience and subjectivity will always be present in the forecast system until more disease/weather algorithms become available. Quality of disease site observations, confidence in the weather forecast data, forecaster experience and model validation are all-important factors in determining how the model will function in its intended use and set the strategy for efficient disease control.