bioRxiv · 10.64898/2026.11.10.765737
Global biodiversity indicators do not fully capture how land cover alters species richness and composition
Abstract
Global biodiversity indicators of land-use impact are increasingly used to track international targets and to inform corporate disclosure, yet few have been tested against independent field data. Because species richness and community composition can change independently, an indicator capturing only one can miss changes in the other. Using standardized bird surveys on three continents, I show that most of seven leading indicators ranked sites within ecoregions no better than the share of natural land cover, for both richness and composition. I developed new estimates of how land cover affects richness and composition, from GBIF occurrence records open to commercial use, and tested them in the same surveys. Only the new composition estimate clearly outperformed natural-cover share for composition, and the new richness estimate predicted field richness as well as an expert-range map in two of three regions. Neither estimate could stand in for the other. Globally, assessments depended on the natural reference: cropland retained about half of the bird species of a forest reference but four-fifths of those of an open natural one. Assessing the state of nature therefore requires an explicitly justified reference, and indicators that report richness and composition as a pair and are tested against independent surveys.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Kadoya, T.. 2026-11-16. Global biodiversity indicators do not fully capture how land cover alters species richness and composition. https://doi.org/10.64898/2026.11.10.765737
Cite the original work for its findings. Save a collection to share your selection of sources.