Transcript Folie 1
Data Market as a Realistic Tool for Decision-Making in IPM Challenge Boll R., Bout A., Doise S., Bernardet X. and Poncet C. Unité de Recherches Intégrées en Horticulture, URIH UR 880, INRA, 400 route des Chappes, 06903 Sophia Antipolis, France. Contacts: [email protected] [email protected] Sophia Data_Market creates an equivalent swap of services for data Acquisition of knowledge regarding pest pressure on cultivated crops leads to invaluable data sets. Data collected from field experiments are abundant and multidimensional in nature. Information is recorded in a warehouse database, analyzed and Data Dimensional Interactive warehouse arrangement Web immediately reported. The On Line Analytical Processing OLAP concept sets dimensional arrangements corresponding to the specific needs of researchers, technicians or advisors. The warehouse database contains - Administrative information on projects and experiments - Material and method parameters - Spatio-temporal biotic and abiotic measures - Distributed climatic recordings - Microsatellite and sequencing pattern catalogues The data_market essentially provides - Protocols for any experimentation - Monitoring module for spatio-temporal reporting with statistics - Management toolbox for field collects and genetic patterns - Calibration and forecasting models - Technical module for setting and running decision rules Learning Programs and Protocols Data Recording Teachware and scoring self-test guides new observer in on-field pest observation. Entry form generator for on-line recordings in order to gain time and reliability on data. Data Management Calibration and Forecasting Models The monitoring mainboard provides spatio-temporal maps and graphics and possibility to update/delete/extract data Models for the decision-making process are continuously improved by entry of new data. Technical Decision Rule Module The data_market characteristics Hardware and software hosted by the INRA platform MySQL, phpMyAdmin and Php languages Database and website are managed by INRA-URIH team Web accessible, private data protected by grants Partial open access and completely free The data_market involved in pest pressure studies Large corpus of various data helps to improve the accuracy of scouting methods by calibration models. Representative sampling grid optimisation is studied by simulation. Dimensional arrangement of data permits the computation for environmental or agronomic indicators as chemical dressing or product quality. Data mining enables the classification determinant factors involved in pest invasion or population outbreaks. These combinations of biotic observations and abiotic measures are tested as predictors for the forecasting models. Application for setting and running technical decision rules for plant protection. The data_market involved in plant protection IPM programs aim to decrease the use of pesticides. This challenge depends mainly on pest pressure estimations and on risk-taking behaviors of the growers. Fields observations performed by weekly scoutings are recorded into the database and immediately capitalized. The monitoring panel reports evaluation and prediction of the main pest pressures. The objective is to help the crop manager to make decision with the most reasonable and efficient plant protection. We develop a software to formalize and use rules for technical decision.