Recognition and completeness: two key metrics for judging the utility of citizen science data

Abstract

Biodiversity citizen science data are being collected at unprecedented scales, and are key for informing conservation and research. Species‐level data typically provide the most valuable information, but recognition of specimens to species level from photographs varies among taxa. We examined a large dataset of Australian photographic observations of terrestrial invertebrates uploaded to iNaturalist to quantify recognition to species across different taxa. We also quantified the proportion of Australian species that have been uploaded to iNaturalist. Across 1,013,171 observations covering 14,663 species (17.8% completeness), 617,045 (60.9%) were recognized to species. Dragonflies/damselflies and butterflies were the best‐recognized and most complete taxa, and therefore represent the best groups for researchers and managers intending to use existing iNaturalist data at large spatial and temporal scales. The recruitment of additional experts to identify records, and enhanced support for accessible resources for hard‐to‐identify taxa, will likely increase recognition for other taxa.