Web-Based Decision Support for Sustainable Pest Management in Fruit Orchards: Development of the Swiss System SOPRA

Abstract

Plant protection in modern orchard management relies on precise timing of monitoring and control of pest populations. Particular stages within the life cycle of insects and other pests have to be surveyed throughout the season by different monitoring procedures in order to establish economic injury levels and to test for the need of control measures, respectively (Kogan, 1998). Growers need precise information about when these stages are present and corresponding techniques have to be applied or monitoring has to be carried out (Fig. 1). Especially in sustainable production applying integrated pest management (IPM) the pests necessarily have to be monitored in order to assess the risk and consequently to judge the requirement of an intervention with a control measure (Norton & Mumford, 1993; Dent, 1995; Pedigo & Rice, 2005). Precise timing of monitoring measures assures reliability of the results and saves time during the decision finding process (Fig. 1). In case control is needed, all regarding actions taken need to be scheduled in relation to crop and pest phenology. Modern control measures even rely on more precise timing especially when modes of action aim at very specific developmental stages of the pest (Blommers, 1994). Besides increasing efficacy, the precise timing of plant protection measures reduces their side effects and may substantially reduce the number of treatments and thereby resources and money spent during the process. The required knowledge on the phenology of the pest populations can be established by forecasting systems that at the best are connected with information on the pests and possible management options to decision support systems. Hitherto, temperature sums and more recently simulation models have been used in tree fruit growing to predict the phenology of pests and hence to facilitate timing of monitoring and plant protection measures. However, aside from a few exceptions (e.g. Welch, et al. 1978; Morgan & Solomon, 1996), the simulation models to predict the phenology of fruit pests are often not designed to be used by growers, consultants or extension services. Also they are often based on very different approaches and programming languages or require special driving variables, which makes them even difficult to use by extension services (Rossing et al., 1999; van der Werf et al., 1999).