Missouri Botanical Garden's Center for Botanical Informatics is working with the Flora of North America (FNA) to create the first fully electronic floristic research project. The result will be an ever-expanding, continually refined digital library containing scientifically authoritative, up-to-date information on the approximately 20,000 species of vascular plants and bryophytes of North America north of Mexico. The library will contain a rich mix of documents, maps, illustrations, computational tools, library services, and, perhaps most important, FNA's database. With over 750 contributing scientists, FNA is one of the largest collaborative research projects ever assembled. The project began several years ago as a paper-based publishing effort; however, traditional publishing methods have not scaled to the task, and FNA has had difficulties meeting its ambitious production schedule. It has become necessary to move FNA into an electronic framework. Doing so is forcing us to confront difficult challenges that lie at the intersection of electronic publishing, large-scale scientific database activities, Internetbased project coordination, and digital libraries.
In order to improve FNA's rate of production, it will be necessary to rethink our fundamental approach to scientific database activities and scientific publishing. Specifically, it will be necessary to reduce the algorithmic complexity of FNA's 'database publishing' task. We propose to accomplish such a reduction by developing a 'common form' coordination system in which all publishing activities occur over the common substrate of a single, virtual database representation of the core information. The strategy is to change the character and number of rate-limiting, interpersonal dependency relationships that exist in FNA's work processes. Within this framework, concepts such as 'manuscript' and 'volume' become abstractions over the base information, and edit and review consist of interactions between people and the common form which can be managed by the coordination system. This work has important implications for other large-scale, information-intensive scientific enterprises. This paper presents an analysis of FNA work processes and an overview of how we are approaching the construction of this important digital library.