Biodiversity conservation is vital for sustainable development, poverty alleviation, and a livable planet. It sustains ecosystems that support human livelihoods, particularly in developing nations, by providing essential resources like food, clean water, and climate stability. Healthy ecosystems are crucial for agriculture, fishing, and forestry, which reduce poverty and foster sustainable economic growth. Additionally, biodiversity fuels industries such as ecotourism and pharmaceuticals, generating jobs and income. Maintaining biodiversity strengthens ecosystem resilience, stabilizes climate, and mitigates climate change impacts, making it key to environmental health and global sustainability.
But, the world is losing biodiversity at an alarming rate. Despite growing awareness of our reliance on nature, the planet faces a severe biodiversity crisis, with one million species nearing extinction and extinction rates 1,000 times above the natural rate. A IPBES report highlights the urgent need for action, echoing findings from the Living Planet Index, which shows a 68% decline in global biodiversity since 1970.
Historically, limited public data on key ecosystems has hindered conservation efforts. The lack of accessible data on vital ecosystems and habitats has slowed progress, particularly in developing regions. Traditional metrics focus on terrestrial vertebrates, while emerging threats outpace the slow update cycles of information managed by underfunded institutions. This data gap impedes effective conservation strategies.
Effective conservation requires current, location-specific data to balance development goals. Advances in data science have expanded access to high-resolution spatial and temporal data, transforming our understanding of biodiversity. Improved computational techniques now allow vast geospatial data from satellites, drones, and sensors to be processed and analyzed at unprecedented scales. Integrating machine learning and AI enhances prediction accuracy and insights, supporting precise monitoring and informed decision-making, which align conservation with development goals.
The World Bank produces essential datasets that drive sustainable development efforts, as it relies on data for policymaking, program implementation, and impact evaluation. It produces a wide range of datasets across critical areas like infrastructure, environment, and economics, essential for guiding sustainable growth. These detailed datasets enable targeted policy interventions, and their adherence to stringent quality standards ensures reliability.
Our datasets on biodiversity offer valuable examples. In 2022, we integrated plant data from Borgelt et al. (2022), ant data from Kass et al. (2022), and data on amphibians, fish, mammals, and reptiles from the IUCN, as well as birds from BirdLife International to produce a database of habitat maps for over 90,000 species. That database includes habitat maps for more than 38,000 endemic (unique to a country) species and 2,000 small-occurrence region (areas less than 25 km²) species. We assessed extinction risks for over 75,000 species, primarily using the IUCN Red List.
Recently, we expanded our dataset with millions of georeferenced reports from the Global Biodiversity Information Facility (GBIF). Using machine-based pattern recognition, we generated occurrence maps from georeferenced species reports. We validated these occurrence maps against expert-produced maps for mammals, ants, and vascular plants, confirming close similarities in global distribution patterns.
The resulting Global Biodiversity Species Global Grid database includes over 600,000 species, including arthropods, mollusks, plants and fungi, across terrestrial, freshwater, and marine environments, alongside more traditionally-represented vertebrates (amphibians, birds, fish, reptiles, and mammals). The figure below shows the composition of our dataset. This expanded database reveals global distribution patterns, enhancing conservation planning.
Total count of species in our database
Effective conservation plans depend on accurate data about species distribution and threats. Using our new occurrence maps, we identified 272,189 endemic species and 85,310 species under threat due to limited occurrence regions. See the bar charts provided below.
A. Total Count of Endemic (unique to a country) Species
B. Total Count of Small-Ocurrence Region (areas less than 25km) Species
While the IUCN provides threat categorizations for some species, the GBIF data now greatly outnumber those previously assessed for extinction risk. To fill this gap, we analyzed location-specific threats and protection measures, estimating extinction risks for 512,675 GBIF species not yet assessed by the IUCN. The Box-Whisker plot below shows the spread of estimated extinction threats across species groups.
Estimated Extinction Threats of all GBIF Species
Our results show that expanding species representation widens the scope of biodiversity conservation, uncovering many more potentially threatened species worldwide and significantly altering "conservation hotspot" maps.
Our datasets will support the Global 30x30 biodiversity initiative and more. The Global 30x30 initiative, part of the Kunming-Montreal Global Biodiversity Framework, aims to protect 30% of the world’s terrestrial and marine areas by 2030. Our datasets will help policymakers assess potential 30x30 sites by considering biodiversity, extinction risks, and economic factors. They will also aid in Environmental Impact Assessments (EIA) and support the World Bank's implementation of ESS6 (Biodiversity Conservation) during project planning and execution.
In conclusion, the GBIF's expanding global species database is a valuable resource for conservation. Our estimation algorithm is designed for rapid updates, supporting the ongoing growth of open-source GBIF reports and enabling broader applications in biodiversity conservation.
If you’re interested in the research on Species’ Occurrence Regions, please see Revisiting Global Biodiversity: A Spatial Analysis of Species Occurrence Data from the Global Biodiversity Information Facility, World Bank Policy Research Working Paper 10821.
For more on estimation of extinction threats, visit Estimating Extinction Risks with Species Occurrence Data from the Global Biodiversity Information Facility. World Bank Policy Research Working Paper 10822.
The authors gratefully acknowledge financial support from the Global Environment Facility.
Join the Conversation