Search

Problem: Often spreadsheets are used as pseudo-databases for the storage of plot-based survey data, but they have major limitations in scalability, concurrent access and data retrieval. Paper-based surveys require timeconsuming data entry. They contain potential inconsistencies (e.g. miss-spellings, abbreviations, missing values), particularly if coming from different observers due to unenforceable data standards. Methods: We analysed more than 30 years of data collected in the Northern Jarrah Forest (NJF) of south-western Australia, comprising c. 31,000 plots (c. 550,000 species records) and associated environmental variables stored across multiple spreadsheets in the development of our free and open source floristic information management system (FIMS). Data dictionaries were developed for each spreadsheet before being combined into a unified standard. OpenRefine software was used to ensure adherence to the standard, including correcting inconsistent field order in different files, removal of redundant or irrelevant fields, abolishing synonyms and abbreviations, and deleting incomplete rows. Database design and normalisation rules ensured the removal of repeating groups and the provision of fields for each retained attribute. Geometry was stored using spatial objects available in PostGIS whilst maintaining an otherwise relational database using PostgreSQL. Results: FIMS provides a spatial database system for storing, accessing and retrieving floristic survey data. FIMS includes a mobile data collection module for use on tablet technology with autonomous database synchronisation and one-step data entry to eliminate transcription and associated errors. Spatial data types enable the retrieval of data for viewing and analysis within most Geographic Information Systems and statistical software. It promotes portability and adaption to other locations and studies via the provision of all necessary code. [ABSTRACT FROM AUTHOR]