Gaps and biases in the ecological evidence underlying global biodiversity impact assessments

Vilma Sandström , Maria Hällfors, Sami El Geneidy, Aapo Kahilainen, Janne Kotiaho, Essi Pykäläinen, Jani Salminen, Laura Scherer, Päivi Sirkiä, Veera Norros

1. Halting global biodiversity loss requires methods that link biodiversity impacts to both their direct drivers and the human activities that generate them. Biodiversity footprint assessment has emerged as a key framework for establishing these connections, resulting in the rapid development of diverse methods and applications. Central to these approaches are characterisation factors (CFs), which translate environmental pressures, such as land use, into quantified biodiversity impacts. CFs are based on empirical ecological studies that vary in taxonomic, geographic and methodological scope, potentially introducing uncertainty, bias and limited representativeness into biodiversity footprint assessments.
2. We systematically examined the empirical foundations of land-use pressure–impact relationships underpinning CFs in four widely used global biodiversity impact assessment methods. We traced the original empirical studies on which the CFs were based, compared overlap among the methods’ publication bases, and assessed their taxonomic, geographic representativeness. We also compared these studies with a recent global database of human impacts on biodiversity and identified priority ecoregions for future research by combining evidence gaps with the occurrence of Key Biodiversity Areas and high anthropogenic land-use pressure.
3. The four methods relied on markedly different but partly overlapping evidence bases. Across methods, the underlying evidence showed pronounced taxonomic and geographic biases. Vertebrates and vascular plants were generally overrepresented, whereas invertebrates, amphibians, reptiles, fungi and many other groups were poorly represented or absent, although the extent of these gaps varied among methods. Evidence was concentrated in (sub)tropical moist broadleaf forests and in Asia, Europe and the Americas, with substantial gaps in Africa, Australia and Oceania.
4. This uneven representation has important implications for the interpretation, robustness and global comparability of biodiversity footprint assessments. Building on our findings, we identify priority regions for future research and highlight opportunities to strengthen biodiversity impact assessment by incorporating more balanced ecological evidence into CF development and promoting structured, assessment-relevant biodiversity data collection. Greater transparency about the empirical foundations, representativeness and limitations of CFs is also essential to support more robust biodiversity footprint assessments and better-informed strategies for mitigating biodiversity loss.