The million of specimens contained in biological collections (natural history museums and herbarial) represent extensive, high-quality documentation of the Earth's biota. Collection data are stored, however, in different data structures under different database management systems, and are consequently different to synthesize. In this project, museum data specialists computer scientists are collaborating to build a federation query server from a web-accessible schema library and an extensive suite of scientific management tools (developed by the OPM Project, Lawrence Berkeley National Laboratories). The schema library documents the structures of individual collection databases and a reference model of shared concepts. The data management tools use inter-schema mappings to transform queries and data between the reference model and local databases. Biological collections data can be used to determine species distributions; multi-species patterns of distribution and diversity; and how distributions are influenced by environmental factors (e.g., climate) and land-use practices. This project will make collections data more accessible by automating the repetition and complexity of querying different collection databases, and will help deliver collection data into the emerging National Biological Information Infrastructure. The query server will complement the analysis, visualization, and prediction tools that are being developed at the San Diego Supercomputing Center (for the US-Organization for Biodiversity and our understanding of global climate change. This work is supported by the Database Activities program in the BIO directorate.
