Ecologists struggling to understand rapidly changing environments and evolving ecosystem threats need quick access to relevant research and documentation of natural systems. The advent of semantic and aggregation searching (e.g., Google Scholar, Web of Science) has made it easier to find useful literature across disciplines and publishers. However, locating information about a specific ecosystem, species or landscape remains hindered by current search technologies. The science community still relies on concepts of publication cataloging and searching dating back to the late 1800’s (e.g., keyword, topic, text, and author searching). The vast majority of published research conducted on ecosystems around the world is tied to specific places, and these locations can be exploited to search for literature based on geography in addition to traditional terms. Our objective was to develop an integrated geographic and thematic (i.e., geo-semantic) system for searching for published ecological literature. We parsed article text from back issues of ecological journals to extract reported study area locations (place names as well as geographic coordinates). Here we look at only papers reporting geographic coordinates. Citation information, keywords, abstracts, and standardized geographic coordinates were added to a database. A web interface was created to search for literature based on geo-semantic queries.
Results/Conclusions
Results are available for the last six years of two journals, Rangeland Ecology and Management (REM) and Journal of Arid Environments (JAE). Additional journals are being added to the database. Of the 2,335 studies published in REM and JAE from 2005 to 2011, 76.5% of studies reporting a study area included geographic coordinates; the rest included only place names. Of the 1,721 REM and JAE studies reporting geographic coordinates, eight different coordinate formats were used. Obvious location errors were found in 36 studies (2.1% of studies reporting geographic coordinates). Beyond easily seeing what has been published for specific areas, georeferenced publication databases open new possibilities for discovering and using ecological knowledge. Users can analyze distributions of published knowledge about a species or topic, or identify relevant knowledge sources from similar areas based on themes like soils and climate. To promote robust geo-semantic literature searching in the future we recommend: 1) adopting standard geographic coordinate formats in journals to increase ease of location extraction and validate coordinates, 2) requiring authors to report geographic coordinates for studies unless privacy or proprietary concerns dictate otherwise, and 3) making geographic information publically available as a part of basic article citation metadata.