Beyond names: the Berlin Model for taxonomic information processing
Berendsohn, Walter. Botanic Garden and Botanical Museum Berlin Dahlem, FUB Department of Biodiversity Informatics and Laboratories, Königin-Luise-Str 6-8, D - 14191 BerlinThe Berlin Model, a comprehensive relational information model for taxonomy based on the earlier published IOPI Model has been implemented and is currently in use by several botanical projects.
There is a need to go beyond names if an index system for biological content is to be created. Names stand for scientific concepts of a taxon's circumscription. Increased knowledge may lead to new concepts, and opinions about a taxon concept may vary as to its content as well as to its classification. Data accumulated for German plants show that literature in current use different concepts in up to 40% of the taxa treated. Due to classification differences, nomenclatural stability is even lower. This practically disqualifies scientific names as unique keys to organism groups in a distributed information network.
Several taxonomic systems support concept related data storage in some way (e.g. by allowing to include misapplied names or by giving lists of qualified synonyms). In the Berlin Model concepts are expressed as a name in the sense of a specific reference. These "potential taxa" form the heart of the system. Our approach to software development allows to share data definitions, database procedures and functions, parts of the business logic and access and editing software across several projects.
The model's taxonomic core comprises modules for names, taxonomic concepts, references, and uncomplicated factual data. The AlgaTerra Project (see Glück & al.) has elaborated a comprehensive structural extension of the name module for typification information. This extension will be added to the core once practical testing by AlgaTerra is concluded.
The generic module for factual data allows for simple textual facts, which are linked to taxonomic concepts but can be individually referenced. This standard module is used e.g. by the Dendroflora of El Salvador project for statements of plant uses and habitat distribution of the species in the country.
Project specific model extensions can be tailored according to a project's specific needs. For example, distribution and occurrence data require more elaborate but differing schemes in the Euro+Med (Flora Europaea), the Med Checklist, and the IOPI databases. For IOPI data, TDWG standards are implemented. AlgaTerra is working on an extension for molecular data, and for the Dendroflora of El Salvador, a specimen citation extension is designed.
The MoReTax-2 project is working on an extension that will allow tuning the rules regarding transmission of factual information from one potential taxon to another. This "transmission engine" will be prototyped for the German Flora. Already existing concept-based checklist datasets will be used to connect data from various sources into a common information system at the Federal Agency of Nature Protection. The German GBIF Node for Botany plans to use that database for query expansion and qualified response in the GBIF access to data about German biodiversity.
An important design principle of the model and the Berlin Model based software is flexibility regarding user demands. Variable atomization allows to preliminary treat unstandardized data. You can, but you don't have to use a concept based approach, the model fully supports e. g. conventional checklist writing. Easy to use tools to support the editing processes and produce traditional output are provided and will be demonstrated (see Döring et al.).