weeDive: Making Long-Term Arable Vegetation and Monitoring Data FAIR

www.julius-kuehn.de Johanna Bensch1,2*, Christoph von Redwitz2, Bärbel Gerowitt1 Case Studies: Testing Our Blueprint weeDive: Making Long-Term Arable Vegetation and Monitoring Data FAIR Johanna Bensch1,2*, Christoph von Redwitz2, Bärbel Gerowitt1 •Agriculture runs on divers data –from field notes to climate records –each with its own quirks and needs •To unlock their full potential, we must first understand these needs before data can deliver real value. •We draft a blueprint with challenges, solutions, and benefits for working smater with agricultural data using three case studies. Turning agricultural data into actionable knowledge Challenge Solution Heterogeneous data Harmonization through standardized data structurs Different scales and time points; varying levels of detail Alingment of scales, clear seperation of raw and derived data Inconsistent plant names Use of controlled vocabularies and metadata Details of collection person-specific, staff turnover Keywording and metadata Missing documentation of surveys Detailed annotation of methods and time points Data often designed for a single research question Standardized formats so data can be integrated across studies •Publication of monitoring data Monitoring data was integrated into an existing database following our blueprint — straightforward and effective. •Resurrection of a 19 years long crop rotation experiment Different levels of data quality need to be harmonized and published in a comprehensive way. In this case the data entails, different randomization approaches, different frequencies in data generation, changing areas, specific crop rotations. This results a various challenges. Applying the blueprint here is more complex but achievable. •Establishing of database for biodiversity connections Links between arable weeds, birds, and insects are derived from literature. This needs to be published in a flexible and easily updateable way. Blueprint use still under development — challenges remain on how to publish while keeping it expandable. The development of an exemplary FAIR workflow as a blueprint shows so far: •FAIR preparation of divers data with domain-specific data is complex •Unlocking the added value of highly complex data on arable vegetation requires specific structures •Long-term experiments with agro-ecological and organismic focuses have specific requirements 1University of Rostock, Faculty of Agriculture, Civil and Environmental Engineering, Rostock 2Julius Kühn-Institute (JKI) –Institute for Plant Protection in Field Crops and Grassland, Braunschweig *[email protected] Process of FAIR (Findable, Accessible, Interoperable, Reusable) preparation and provision of a specific, existing dataset