Methods: Selecting Sample Areas and Measuring Global Conservation Coverage

Selection of the Sample Area

Our analysis built on a previous study⁸ that identified 10% of the terrestrial realm for measuring conservation area coverage and ecological representativeness. The approach minimized the number of selected countries and territories to reduce subsequent data-collection requirements (Extended Data Fig. 1a and Supplementary Discussion).

The process began by identifying 10 factors that influence conservation area extent and global biodiversity patterns:

1. Biomes
2. Elevation
3. Government effectiveness
4. Income per capita
5. Islands and continents
6. Land cover
7. Latitude
8. Population density
9. Realms
10. Subregions

Each factor was subdivided into several “features” (Supplementary Table 1). For categorical factors, each category was treated as a separate feature. For example, income per capita was divided into the World Bank categories of low income, lower-middle income, upper-middle income and high income⁸. Continuous factors were divided according to appropriate break points in the data. Government effectiveness, for example, was divided into four features using World Bank scores of 0–25, 25–50, 50–75 and 75–100⁸.

This process produced 89 features, which were mapped using the best available datasets⁸.

Marxan Sample Selection

The previous study used Marxan software to identify a sample covering 10% of the global extent of each of the 89 features⁸. This ensured that the influence of the different factors was proportional to their representation at the global scale.

The 10% coverage target was selected through a sensitivity analysis that balanced the total area selected against the number of countries and territories included in the sample⁸. Marxan uses a simulated annealing algorithm, running the software multiple times and producing a near-optimal output from each run.

The Marxan analysis was conducted in two stages, with 1,000 runs in each stage (Supplementary Methods).

Stage 1: Identifying Countries and Territories

Stage 1 treated each country or territory as an analysis unit. Countries with land areas greater than 1,000,000 km² were divided into units based on their highest subnational administrative level, such as states or provinces.

This stage identified the best of the 1,000 outputs: the result containing the fewest countries and territories needed to meet all 10% global coverage targets (Fig. 1a).

Stage 2: Selecting Sampling Units

Stage 2 removed superfluous land from the countries and territories selected in Stage 1. This reduced coverage of each feature to as close to 10% as possible and balanced the influence of the features.

The Stage 1 countries and territories were divided into 100 km × 100 km sampling units. Marxan was then used to identify near-optimal sets of sampling units that met all 10% global coverage targets while minimizing the total area selected.

For this study, we used the 1,000 outputs from the Stage 2 analysis (Extended Data Fig. 2). This allowed us to calculate conservation area coverage for each set of sampling units and estimate associated uncertainty.

Conservation Area Dataset

We collected available conservation area data for our focal sample of 25 countries and territories from multiple sources.

The World Database on Protected and Conserved Areas (WDPCA) is the official dataset for measuring global protected-area coverage². It was launched in 2025 by combining the pre-existing World Database on Protected Areas (WDPA) and World Database on OECMs (WD-OECM).

We downloaded protected-area boundary data for each country and territory in our sample area in May 2022. No OECMs were listed in the sample area by that date, so the data were extracted from the WDPA.

We followed the recommended method for calculating protected-area coverage⁵¹. We removed UNESCO Man and the Biosphere Reserves (MAB), as well as areas with a status of “proposed” or “not reported”.

The protected-area spatial data included:

• 36,835 boundary polygons
• 138 point locations with an areal extent
• 1,077 point locations without an areal extent

For point data with a reported areal extent, we created a circular buffer around each point corresponding to the reported area. Point locations without area data were removed. We then combined the buffered point data and polygon data into a single spatial data layer for each country and territory.

Including UNESCO Man and the Biosphere Reserves

MAB sites are excluded from WDPA coverage calculations because the database does not distinguish between their core, buffer and transition zones⁵¹. To include MAB sites in our analyses, we mapped them using information available from the UNESCO website⁵², which provides point coordinates and the size of the core, buffer and transition zones.

For MAB sites with suitable data, we created a circular buffer around the point coordinates to represent the listed core-zone area. We excluded buffer and transition zones because expert feedback indicated that these zones did not meet the criteria for the conservation area categories defined earlier (Box 1).

National and Regional Conservation Area Data

We also downloaded publicly available national and regional conservation area spatial data for each country in the sample (Extended Data Tables 1 and 2). These datasets included national protected-area databases, community conservation areas and other areas that may be managed in ways that conserve biodiversity.

In addition, we asked conservation scientists and practitioners with expertise in conservation area types and designations in the sample countries to provide any other relevant spatial datasets. Our aim was to base the analysis on the most complete available set of conservation area data.

Some of the information was politically sensitive, particularly data on conservation area boundaries. These datasets were shared on the condition that we would not publish the raw spatial data, summarize results at the national level or make the datasets publicly available. Expert input during the later stages of the analysis was also anonymized.

Ethics approval for these processes was granted by the University of Kent Faculty of Social Sciences Research Ethics Advisory Group for Human Participants (ID 20221648467987119).

Overall, the analysis used:

• Two publicly available global datasets
• 14 publicly available national datasets from five countries
• 51 non-public national datasets from 16 countries and territories

Categorizing the Conservation Area Dataset

Our analysis focused on protected areas (PAs) and other effective area-based conservation measures (OECMs). We developed three conservation area categories to describe them (Box 1), using the International Union for Conservation of Nature (IUCN) and Convention on Biological Diversity (CBD) definitions that underpin progress measurements for Global Biodiversity Framework (GBF) Target 3 (Supplementary Methods), together with the metadata of the spatial datasets collected.

Each conservation area was classified into one of the three categories. Areas that did not meet the criteria for any category were removed from the dataset.

We first produced a provisional classification for each area based on its metadata. These provisional assessments were then provided through an online form to local experts in each sample country, who were asked to check, correct and comment on the classifications.

When experts disagreed about the appropriate category, we used the majority response. In the event of a tie, a working group of six co-authors made the final decision. A sensitivity analysis tested the effect of this decision-making process and found that the results were robust to change (Supplementary Methods).

In general, conservation areas with the same national designation were assigned to the same category. However, experts identified exceptions and used the available data to assign individual sites to different categories where appropriate (Supplementary Methods).

Category 1: Globally Recognized Protected Areas

Category 1 includes all conservation areas included in the WDPCA with a status of “inscribed”, “adopted”, “designated” or “established”². These are protected areas that meet the criteria specified in the recommended method for calculating protected-area coverage⁵¹.

Category 2: Nationally Recognized Protected Areas

Category 2 includes sites that are not in the publicly available WDPCA but are included in national databases and meet the IUCN definition of a protected area. Relevant experts assessed these sites using available metadata, peer-reviewed evidence and local knowledge.

Sites may be absent from the WDPCA for several reasons:

1. Delays in national agencies submitting updates about new sites⁵³
2. Restrictions placed by countries on the public availability of their data²
3. National legislation that does not recognize particular protected-area governance types, including privately owned and governed sites^7,22

Category 3: Possible OECMs

Category 3 identifies areas that could potentially be counted as OECMs in the future. Designating an area as an OECM and listing it in the WDPCA requires an in-depth assessment of its management, governance and biodiversity values. It also requires consent from the group governing the site²⁶.

Our definition of possible OECMs was based on translating the IUCN site-level tool for identifying OECMs⁵⁴ into six components that could be assessed by national experts:

1. Not a protected area. Sites that could be classified as globally or nationally recognized protected areas were excluded.
2. A geographically defined area with agreed and demarcated boundaries. We used only spatial datasets with demarcated boundaries produced by relevant government agencies or by groups responsible for governing, managing or supporting the sites.
3. Governed by a specified entity or group. Relevant experts assessed this criterion using available metadata, peer-reviewed evidence and local knowledge.
4. Governance likely to continue indefinitely. Relevant experts assessed the likely durability of governance using available metadata, peer-reviewed evidence and local knowledge.
5. Managed in a way that contributes to in situ conservation. This criterion applied regardless of whether conservation was the stated intention of the management regime.
6. Contains important biodiversity values. Relevant experts assessed this criterion using available metadata, peer-reviewed evidence and local knowledge.

We used a broad interpretation of “important biodiversity values”. Although the phrase appears in the IUCN site-level identification tool⁵⁴, the OECM definition specifies only that sites should have “long-term outcomes for the in situ conservation of biodiversity”. We therefore retained all sites that included natural green space, even when the site was designated for another purpose, such as protecting built heritage.

Spatial Analyses

Preparing the Data for Analysis

All analyses and data processing were conducted in R v.4.2.1⁵⁵ and QGIS v.3.16.3⁵⁶. Unless otherwise stated, all datasets were reprojected to the Mollweide equal-area projection and analyzed in that projection.

To prevent double counting, we used the st_difference function in the R sf package⁵⁷ to remove duplicates and overlaps among datasets.

When a polygon appeared in multiple datasets, or when overlapping polygons belonged to more than one conservation area type, the overlapping area was assigned to the lowest-numbered type (Box 1). For example, if a globally recognized protected area overlapped with a possible OECM, the overlapping area was categorized as a globally recognized protected area. This ordering reflects differences in recognition and permanence.

Conservation Area Coverage Statistics

Using the combined conservation area dataset, we calculated the extent of each conservation area type in every 100 km × 100 km sampling unit. We then used these results to determine the percentage coverage of each conservation area type for each of the 1,000 sampling-unit sets⁸ (Extended Data Fig. 2).

The mean coverage of globally recognized protected areas across the sets was 16.15%, slightly higher than the 15.79% global coverage reported by the WDPCA in May 2022 (ref. 27).

Because debates around the 30 × 30 target are contentious, we adjusted the results to account for the slightly higher globally recognized protected-area coverage in our sample. For each of the 1,000 sets, we divided the 15.79% WDPCA value by the percentage coverage of globally recognized protected areas in that set. We then multiplied the conservation area extent results by this adjustment value.

This adjustment ensured that the percentage coverage of globally recognized protected areas matched the reported global WDPCA coverage. The other conservation area types received the same proportional adjustment (Extended Data Table 3).

We then calculated the adjusted mean percentage coverage and standard deviation for each of the three conservation area types across the 1,000 sampling-unit sets (Extended Data Fig. 1b and Extended Data Table 3).

We assumed that the percentages calculated from the sample were equivalent to those across the entire terrestrial realm. We made this direct extrapolation because Marxan was used to design each set to represent the terrestrial realm by containing 10% of the total extent of every spatial feature known to influence conservation area coverage (Extended Data Fig. 1b, Extended Data Fig. 2, Supplementary Fig. 1 and Supplementary Tables 1 and 2).

Because the 1,000 sets were selected for representativeness, their sampling units are more similar to one another than they would be under random selection (Extended Data Fig. 2). Consequently, the standard deviation values calculated from these sets are relatively low.

Ecosystem Integrity

To assess the current status of ecosystems within the different conservation area types, we used the Ecosystem Integrity Index (EII) dataset²⁸.

The EII measures the structural, functional and compositional integrity of ecosystems. Each 1 km × 1 km pixel receives a value between 0 and 1, based on the lowest integrity score among the three components.

We divided EII values into four quartiles:

• Low: 0–0.25
• Moderate: 0.25–0.5
• High: 0.5–0.75
• Intact: 0.75–1

For each of the 1,000 sampling-unit sets, we intersected the EII data with the conservation area dataset. We then calculated the mean percentage of each conservation area type covered by each EII quartile across all sets.

Agricultural Suitability

To investigate potential pressure from land conversion and possible food-security impacts, we calculated the agricultural suitability of land within each conservation area type across the 1,000 sampling-unit sets.

We used the Global Agricultural Land Resources dataset, which covers 1980–2009²⁹, and grouped values into four categories using the classification system developed by ref. 58:

• Unsuitable: 0
• Marginally suitable: 1–32
• Moderately suitable: 33–74
• Highly suitable: 75–100

As with the ecosystem integrity analysis, we calculated the percentage of each agricultural suitability category within each conservation area type for every sampling-unit set. We then calculated the mean across all sets.

Ecological Representativeness

We assessed the ecological representativeness of the different conservation area types in each of the 1,000 sampling-unit sets by measuring their coverage of biomes and ecoregions.

GBF Target 3 does not specify the amount of each biome or ecoregion that should be conserved to achieve ecological representativeness. This is partly because biodiversity values vary substantially among them⁵⁹. However, biome and ecoregion coverage statistics are widely reported because they indicate how conservation coverage is distributed and can help identify areas with high or low representation⁶⁰.

We did not measure overlap between conservation areas and Key Biodiversity Areas, even though GBF Target 3 refers to conserving “areas of particular importance for biodiversity”^3,61. Key Biodiversity Areas are generally assessed at the national level, and some countries have not yet assessed their sites for a range of taxonomic groups using the revised Global Standard⁶². Therefore, although our sample was selected to represent broad biogeographic patterns, it may not represent patterns of Key Biodiversity Area assessment.

Ecoregion and Biome Coverage

Our analyses used the Terrestrial Ecoregions of the World dataset³³. For each of the 1,000 sampling-unit sets, we calculated the total area of each ecoregion and the percentage of that area covered by each conservation area type.

These percentage values were adjusted using the same approach applied to the conservation area coverage estimates. For each set, the ecoregion area in each conservation area type was divided by the percentage coverage of globally recognized protected areas in that set and multiplied by the 15.79% WDPCA value.

We removed lake, rock and ice, and mangrove ecoregions to account for spatial mismatches between country and ecoregion boundaries, following ref. 34.

We used the adjusted ecoregion dataset to calculate the total area of each biome and the total area within each conservation area type. Each ecoregion was assigned to its most relevant terrestrial functional biome (Supplementary Table 5), based on the Global Ecosystem Typology functional biome classification³⁰. This classification was designed to measure progress toward GBF targets.

We then calculated the percentage coverage of each conservation area type within each biome.

Ecoregion Protection Equality

We also used the adjusted ecoregion dataset to assess ecoregion protection equality and coverage in relation to past, current and potential international targets.

First, we removed ecoregions with less than 10% of their total extent in the sample area. This left a median of 173 ecoregions in each sampling-unit set (Supplementary Table 6). We made this adjustment following a sensitivity analysis investigating whether small ecoregion patches that barely overlapped with the sampling units could bias the results (Supplementary Methods).

Such bias could occur because small patches are more likely to be either entirely outside or entirely inside the conservation area network. This can make their apparent percentage conservation coverage very low or very high.

Protection equality is a Gini coefficient-based metric widely used to measure the evenness of protection across ecoregions^34,63. We calculated proportional protection equality across the terrestrial ecoregions of the world³³, using the mean values for the 1,000 sampling-unit sets and the ProtectEqual package in R⁶⁴.

Calculations were conducted iteratively (Extended Data Table 4), beginning with:

1. Globally recognized protected areas only
2. All protected areas, including globally and nationally recognized protected areas
3. All conservation areas, including globally recognized protected areas, nationally recognized protected areas and possible OECMs

Ecoregion Coverage and International Targets

To measure ecoregion coverage against international commitments, we created six percentage-coverage groups:

• 0%: No coverage
• 10%: The 2010 CBD Biodiversity Target, goal 1⁶⁰
• 17%: The 2020 CBD target⁶⁵
• 30%: The 2030 CBD target³
• 50%: Recommendations of the Half-Earth Project⁶⁶
• 100%: Complete coverage

We counted the number of ecoregions in each group as the three conservation area types were added incrementally to the global conservation area network.

These calculations used adjusted ecoregion areas from the sample. As a result, some conservation area coverage estimates may underrepresent or overrepresent global levels. However, because the sample included 173 ecoregions, the results are likely to be indicative of global patterns.

Reporting Summary

Further information about the research design is available in the Nature Portfolio Reporting Summary linked to this article.