BACKGROUND. Data obtained from long-term agro- nomic experiments are a basic starting point for the development, calibration and validation of mathe- matical models of cropping systems. The usefulness of these experiments done in Italy is strongly im- paired by scarce information on easily available da- ta, on their organization and data collecting methods, and by difficulty in accessing them. METHODS. The philosophy and layout of a meta-data- base (ClimagriLT) designed to store and share data re- lated to a set of long-term experiments (treatments, yields, soil and meteorological data etc.) is presented in this paper. The main focus is on data model and management policy of data distribution. The details of the long-term experiments, i.e. locations, treatments, studied factors, data collection methods, availability, distribution, experiment site map are contained in the meta-database. The database stores weather data, soil analysis data, yield and biometrics of crops. The data model was built following the relational database the- ory because it is widely accepted, intuitive and easy to implement. Four steps drove the data model construc- tion: i) requirements; ii) draft of the conceptual struc- ture; iii) draft of the logical structure; iv) normalization. A brief introduction is also given to future applications.
