Open data

Abstract

The availability of data supporting scientific publications is becoming more and more critical to the ecological community. Some ESA journals already require that the primary data supporting an accepted paper be made available, and this may eventually be true for all of them. This change parallels a movement in science generally encouraging – and increasingly insisting – that data be publicly accessible. The move toward open data began in a few disciplines where archiving data became customary, and these fields were joined by NASA, which instituted early policies supporting the free and open exchange of scientific data. While scientists often either accept the notion of open data, or resist it personally for a variety of reasons, the motivation for scientists to provide data in support of their publications is not unidimensional. For ecologists, the desire to test theory against larger data compilations is a powerful motivating factor, as each ecologist's individual studies at a particular site may be representative of a more general principle, or may reflect idiosyncratic site conditions. Ecologists are more and more interested in using larger datasets to complement field studies, or even depending on them as their primary research tool. Transparency is also critical, adding credibility to controversial findings because skeptical or supportive colleagues can inspect the data and draw their own conclusions. This is closely linked to the notion of replication. Many ecological studies are difficult or even impossible to replicate when, for example, the result reflects a particular set of environmental conditions, a sequence of weather events, or a disturbance that is not within the researcher's control. In these cases, the goal of replication may be accomplished by another ecologist repeating the analysis, perhaps using a different theoretical framework, but the same data. In Ecological Applications, papers are often used to support management or policy decisions, and this is not uncommon in the other ESA journals, particularly Frontiers. In such cases, the fact that the data underlying a paper are available for scrutiny can greatly strengthen the credibility of the study. In the most contentious decisions, availability of data can be critical, and unavailable or poorly documented data can have serious consequences. Here, careful archiving and documenting of data and metadata can help prevent serious difficulties and legal complications. There are various challenges associated with open ecological data, some unique to our discipline and others that are common across many fields. Data on endangered species are often shielded to some degree by law, and common sense, and journal editors must accommodate this reasonable restriction. Many ecologists make use of data – collected by management agencies – that are partly protected for privacy and commercial concerns, such as fisheries landings or forest data on private land. Usually, this information can be shared after some degree of processing, or after the original request for raw data is properly modified so as to ensure that confidentiality isn't compromised. In the international realm, some nations share environmental data for research but do not allow redistribution of the raw data, and again, in such cases, additional documentation may be required to allow future replication. Documentation and metadata are crucial. Without adequate description, archived data are just so many bits, and with large, complex datasets, researchers are increasingly taking advantage of technology and informatics to document these datasets. While ecologists use a wide enough range of data, from genomics to geophysics, such that no single standard of documentation applies, many groups are working to define and refine documentation and metadata standards, and applying common sense to questions regarding what a future researcher might need goes a long way. When the analytics are applied to the data, or when the raw data are transformed into an analyzed product, it is as important to archive and document those analytics, and the Society's journals also encourage or require this. Open data is an exciting development for ecology. It opens a number of doors for us as researchers, allowing studies across huge areas, creating more context for field studies, and building credibility with the broader public and with each other. It does add an extra set of requirements, which creates another task for already busy researchers, but at the same time it creates opportunity. Key requirements for open data include appropriate repositories, clear standards, and financial support for the additional effort, and these requirements are increasingly being met. Any author who has concerns about depositing data and analytics in support of an ESA journal paper should contact the relevant Editor-in-Chief. David Schimel Editor-in-Chief Ecological Applications Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA