A high-performance web-based information system for publishing large-scale species range maps in support of biodiversity studies

Functionality, performance and scalability are critical to Web-based information systems for publishing and disseminating large-scale species distribution data. Existing systems do not support dynamic spatial window queries on large-scale species range maps that are important to compute alpha and beta diversities for biodiversity analysis and modeling. In this study, we have developed a main-memory based novel quadtree data structure to represent large-scale species range maps and support dynamic spatial window queries to retrieve a list of species and their area sizes within a query window efficiently. Using the NatureServe's 4000 + bird species range maps, experiment results have shown that the memory footprint of the proposed quadtree data structure representing the range maps of all the species is about 1/6 of the quadtree derived by combining individual quadtrees each representing a species range map. The experiment results have also demonstrated that the query response times of our main-memory spatial database are well below a fraction of a second for query windows as large as 10 × 10°, which are 2–3 orders better than using a typical disk-resident spatial database system.

► Novel main-memory quadtree structure to represent large-scale species range maps.
► Web-based system to query species and area sizes in a dynamically defined window.
► Sub-second response times and 2-3 orders better performance on 4000 + bird datasets.