Taxonomic shortfalls in digitised collections of Australia’s flora

Abstract

Rapid growth in the digitisation of the world’s natural history collections substantially simplifies scientific access to taxonomic and biogeographic information. Despite recent efforts to collate more than two centuries of biodiversity inventories into comprehensive databases, these collections suffer limitations across spatial, temporal and taxonomic dimensions. We assessed taxonomic shortfalls in preserved specimens from 296 plant families native to Australia, for which records have been collated into the Australasian Virtual Herbarium (AVH), specifically addressing the following questions: (1) Based on the number of specimen records per species, which Australian native plant families are under- or over-represented in the collection of preserved specimens digitised in the AVH? (2) To what extent does the distribution of collectors among plant families, or the area occupied by plant families, explain patterns of taxonomic representativeness? We found that the number of preserved specimens per family is not proportional to the family’s known species richness. For 29% of Australia’s plant families (i.e. 86), the number of digitised records constitutes < 50% of the number expected given species richness within those families. Further, only 34% of families (100) have at least 20 specimens digitised for each species recorded in the AVH. Families occupying small areas (< 200 grid cells) are more likely to be under-represented taxonomically, while there is a strong positive correlation between the number of unique collectors and the extent of taxonomic over-representation. A sound understanding of biodiversity is critical for megadiverse countries such as Australia, and identifying biases in digital inventories may help with establishing future sampling and digitisation strategies to enhance taxonomic representation.