A multidimensional assessment of Antarctic terrestrial biological data

Abstract

Aim

The globally significant communities of terrestrial Antarctica face an uncertain future amid growing threats in the region. Emerging data-driven approaches must be leveraged to predict and understand patterns of biodiversity across the continent. A new comprehensive database of Antarctic biological occurrence records, the Biodiversity of Ice-free Antarctica Database, will enable such novel fundamental and applied biodiversity modelling. However, there are limitations of assembled occurrence databases that, if unaccounted for, can result in poor model inference and outcomes. We perform a data quality assessment of the new database to highlight its potential and to identify data limitations that must be considered during modelling.

Location

Antarctica.

Methods

We assessed the coverage of the Biodiversity of Ice-free Antarctica Database across geographic, environmental, taxonomic, and temporal dimensions at several spatial scales.

Results

We demonstrate great potential for the database to improve our understanding of many at risk and poorly known Antarctic functional groups. We also provide evidence for limitations of the database across data dimensions, including low geographic coverage that is biased towards research stations, poor coverage of environmental variation across the landscape, and long periods since records were last collected. The magnitude of these limitations varies substantially by region and spatial scale.

Main Conclusions

In combination, data limitations have a range of implications for terrestrial Antarctic modelling, including heightening the risk of model extrapolation. For future use, we recommend prioritising, mitigating, and presenting context-specific model uncertainty, advancing strategic data collection, and exploiting shared modelling challenges elsewhere in the world to maximise the opportunities for this unique dataset to robustly advance science and conservation in Antarctica.