Contract type: Internship
Duration: 6 months
The successful candidate will join the Agro-Environmental Systems team of the “Environmental Research and Innovation” (ERIN) department. With a team of more than 170 scientists and engineers from life science, environmental science, and IT science, the ERIN department has the necessary interdisciplinary knowledge and skills to tackle major environmental challenges our society is facing today: climate change mitigation, ecosystem resilience, sustainable energy systems, efficient use of renewable resources, environmental pollution prevention and control. The RDI activities of the Agro-Environmental Systems team relate to the development and dissemination of models for reducing pesticide use in cereals, rapeseed and vine based on environmental, agronomic and molecular data. The team operates a pest and disease warning platform for the national agricultural community. The team comprises specialists in phytopathology, entomology, and viticulture able to address the most important crop protection issues in Luxembourg.
In integrated pest management (IPM), pests are controlled, when the costs of control correspond with the damage caused by a pest on a monetary scale, implying that low pest levels are left uncontrolled. The disease incidence at which control actions need to be taken for avoiding economic losses was termed the control threshold. Many models forecasting epidemiologically important parameters such as minimum spore densities needed for infection, infection severity classes or newly expressed symptoms were developed in plant pathology. Unfortunately, none of these models predicts the economically relevant control threshold directly making an application in precision agriculture where pesticides and other inputs shall be used precisely when they are needed where they are needed, difficult. We recently developed a new approach for forecasting when the most damaging fungal pathogen in wheat in Luxembourg needs to be controlled. It is the purpose of this internship to validate this approach with various validation methods including leave-one-out cross-validation and using newly acquired external data.