A standardized workflow for compiling and harmonizing first records of non-native species

Biological invasions are a major driver of biodiversity loss, and tracking non-native species introductions is essential for biodiversity monitoring and policy assessment. First records, defined as the earliest documented evidence of a species occurring in the wild outside its native range, provide a standardized metric to quantify introduction dynamics across regions and taxonomic groups. However, compiling and harmonizing such records from diverse sources remain challenging due to non-standardized terminology and heterogeneous reporting standards. Here, we describe a reproducible workflow for standardizing first records of non-native species from diverse sources. The workflow is implemented in R and is designed to ensure consistency in taxonomy, harmonize geographic locations, convert heterogeneous temporal expressions into four-digit years using explicit transformation rules, and standardize terms following Darwin Core terminology. A key innovation of the workflow is the assignment of a confidence level to each standardized first record, reflecting the uncertainty of the original temporal information. The workflow is open-source, modular, and customizable, allowing users to apply it to their own data or modify standardization rules. It is applied to produce v4.0 of the FirstRecords dataset, which contains nearly 100,000 first records of more than 30,000 taxa and is structured into linked tables covering actual records, taxonomy, locations, and data sources. This represents a 17.8% increase in the amount of available first records compared to the previous version, of which approximately 8% is due to the refinement of the workflow. By providing a transparent and automated approach to data harmonization, this workflow facilitates large-scale analyses of invasion patterns and indicator development, while maintaining data integrity.