Synthesizing disparate datasets for biodiversity studies

How do we synthesize disparate datasets that encompass diverse biomes, taxa, and sampling methods? I describe a general approach and demonstrate its utility to the research workflow of two LTER Synthesis Working Groups that are testing community ecology theory using community survey datasets from terrestrial, freshwater, and marine ecosystems. Our coded workflows integrate online data repositories, cloud storage, version control, and open-source software for statistics and markup. Archived datasets cannot be automatically formatted for analysis. I describe the process and pitfalls involved in coming to understand these data, and highlight the importance of metadata, geolocations, and taxonomic identifiers.