The gap between official and actual coverage stems from a structural blind spot in how the world counts conservation. The headline indicator for Target 3 is derived from the World Database on Protected and Conserved Areas, which merged the long-standing World Database on Protected Areas with the newer World Database on OECMs in 2025. Because most nations only recently began recognizing, mapping and reporting other effective area-based conservation measures, or OECMs, the official statistics are dominated by state-governed protected areas such as national parks and nature reserves. Yet a growing body of national studies has shown that substantially more land is managed for conservation than appears in global tallies, much of it governed by private actors, Indigenous Peoples or local communities. These informally governed lands can, according to earlier research, deliver better conservation outcomes than some state-run protected areas, and they often improve the connectivity and ecological representativeness of conservation networks.

Collecting conservation data from every country would be prohibitively slow, so the team adopted a sampling strategy designed to make a globally representative estimate from just 25 countries and territories. Building on a previous methodological study, the researchers first identified ten factors known to drive both conservation area extent and global biodiversity patterns, including biome type, elevation, government effectiveness, income per capita, island versus mainland status, land cover, latitude, population density, biogeographic realm and subregion. These factors were subdivided into 89 distinct features, and the Marxan conservation planning software, which uses a simulated annealing algorithm, was run in two stages of 1,000 iterations each to identify a sample covering 10 percent of the global extent of every feature. The result was a set of 100-kilometre-by-100-kilometre sampling units distributed across nine whole countries and territories plus subnational sections of 16 more, each of the 1,000 resulting sample sets balancing representativeness against minimal country coverage.

For this sample, the team assembled an unprecedented data haul: two global datasets, 65 national and subnational datasets, of which 14 were publicly available and 51 were shared confidentially by governments and experts. Each conservation area was assigned to one of three categories. The first comprised globally recognized protected areas listed in the World Database on Protected Areas as of May 2022, the same dataset used to inform Global Biodiversity Framework negotiations. The second captured nationally recognized protected areas, sites absent from the global database, whether because of reporting delays, national data restrictions or legislation that fails to recognize certain governance types such as privately owned reserves, but which meet the IUCN definition or national definitions of a protected area. The third category, possible OECMs, identified areas under long-term, specified governance that contribute to in situ biodiversity conservation regardless of primary intent, translated from the IUCN site-level identification tool into six assessable criteria. Local experts reviewed and corrected the provisional classifications, with disagreements resolved by majority response and, where ties persisted, by a six-member working group, a process validated by sensitivity analysis and approved by the University of Kent research ethics committee.

The numbers that emerged are striking. Globally recognized protected area coverage in the sample was 16.15 percent, closely tracking the official 2022 figure of 15.79 percent, which gave the team confidence in the method. To be conservative, all estimates were adjusted by the ratio between sampled and official coverage, a mean correction factor of 0.978. Extrapolating across the terrestrial realm, the researchers calculated that 15.79 percent of land sits in globally recognized protected areas, a further 3.64 percent in nationally recognized protected areas, and 7.38 percent in areas that could qualify as OECMs, for a total of 26.81 percent, with a standard deviation of just 0.239 percentage points. Because most nationally recognized protected areas would likely qualify for international recognition, the analysis suggests the world may be only about 3.2 percent of land area short of the 30 percent target, rather than the 11.5 percent, some 15.5 million square kilometres, implied by official statistics. The true figure may be even higher, since some nations lack the capacity to map their conservation estates at all.

Crucially, the newly counted land is not ecologically trivial. Using the Ecosystem Integrity Index, which scores structural, functional and compositional integrity for every square kilometre of the planet, the team found that globally recognized protected areas, nationally recognized protected areas and possible OECMs all showed similarly high levels of ecological integrity, with formally recognized protected areas holding the largest share of fully intact land. Only 2.5 percent of the land across all three conservation area types was highly suitable for agriculture, based on the Global Agricultural Land Resources dataset, suggesting that expanding recognition would not primarily divert prime farmland. Possible OECMs did, however, contain considerably more moderately suitable agricultural land than protected areas, indicating that many could continue to support livelihoods through low-level use and trade of natural resources while sustaining biodiversity values, a pattern consistent with the stewardship models of Indigenous and community governance.

Yet the study delivers a sobering counterpoint to its own headline. When the researchers examined ecological representativeness, the picture darkened considerably. Coverage of the six terrestrial functional biomes, classified according to the Global Ecosystem Typology designed for Global Biodiversity Framework monitoring, remains deeply skewed, with the polar and alpine biome enjoying by far the highest protection, a well-documented artifact of the low economic opportunity costs of conserving cold, remote land. Possible OECMs play a particularly important role for tropical and subtropical forests and for savannas and grasslands, but they largely overlap with ecoregions already well represented in formal protected area networks. The protection equality metric, a Gini-coefficient-based measure scoring 1 for perfectly even coverage across ecoregions, stood at a dismal 0.405 for globally recognized protected areas and rose only to 0.505 when all conservation area types were included. Of the median 173 ecoregions in each sample, 80 had less than 10 percent coverage in globally recognized protected areas; adding every other conservation area type reduced that number only to 51. Meanwhile, the count of ecoregions meeting the 30 percent threshold rose from 43 to just 80.

This disconnect between quantity and quality explains a paradox that has long troubled conservation scientists: alarming biodiversity loss continues even as conserved area grows. The extra land identified in this study tends to sit in ecoregions already well covered, leaving many others, particularly those with high economic value for agriculture and development, under-represented. The authors argue that simply hitting the area component of Target 3 will therefore do little to halt ongoing losses unless governments simultaneously pursue the target’s other components: effective management, equitable governance, ecological representativeness and connectivity. Future expansion, they contend, must be guided by systematic conservation planning that accounts for trade-offs with food security and sustainable trade, respects the cultural significance of land for local peoples, and targets scarce resources toward priority gaps rather than replicating protection in already secure landscapes.

The findings carry profound implications for the politics of 30×30. Critics have warned that near-doubling of conserved area could harm vulnerable communities and entrench top-down, donor-driven exclusion, while spreading conservation funding ever thinner across an under-resourced estate. The new estimate reframes that debate: if more than a quarter of the terrestrial realm is already managed in ways consistent with protected area and OECM definitions, the path to 30 percent may lie less in acquiring new land and more in formally recognizing and empowering the people already stewarding it. That recognition, the authors stress, hinges on free, prior and informed consent from the Indigenous Peoples and local communities who govern much of this land, many of whom may be wary of official designations and their potential restrictions. Governments will need to build effective identification, implementation and funding mechanisms, backed by sustained donor investment, and to strengthen global reporting capacity so that national datasets, marine realms included, can complement the official databases. Measuring progress honestly, the study concludes, is the first step toward a conservation network that is not just large, but effective, representative and equitable, the kind of network on which the world’s remaining biodiversity ultimately depends.

Subject of Research: Global terrestrial conservation area coverage and progress toward the Kunming-Montreal Global Biodiversity Framework 30×30 target

Article Title: Global conservation area coverage substantially exceeds current estimates

Article References: Global conservation area coverage substantially exceeds current estimates. (n.d.). https://doi.org/10.1038/s41586-026-11157-4

Image Credits: AI Generated

DOI: 10.1038/s41586-026-11157-4

Keywords: 30×30, protected areas, OECMs, biodiversity, Global Biodiversity Framework, conservation coverage, ecological representativeness, Kunming-Montreal, Indigenous lands, protected area databases, Convention on Biological Diversity, conservation planning