A Global Deal For Nature: Guiding principles, milestones, and targets

See allHide authors and affiliations

Science Advances  19 Apr 2019:
Vol. 5, no. 4, eaaw2869
DOI: 10.1126/sciadv.aaw2869


The Global Deal for Nature (GDN) is a time-bound, science-driven plan to save the diversity and abundance of life on Earth. Pairing the GDN and the Paris Climate Agreement would avoid catastrophic climate change, conserve species, and secure essential ecosystem services. New findings give urgency to this union: Less than half of the terrestrial realm is intact, yet conserving all native ecosystems—coupled with energy transition measures—will be required to remain below a 1.5°C rise in average global temperature. The GDN targets 30% of Earth to be formally protected and an additional 20% designated as climate stabilization areas, by 2030, to stay below 1.5°C. We highlight the 67% of terrestrial ecoregions that can meet 30% protection, thereby reducing extinction threats and carbon emissions from natural reservoirs. Freshwater and marine targets included here extend the GDN to all realms and provide a pathway to ensuring a more livable biosphere.


Nature conservation efforts, like climate change policies, are being reassessed in the midst of a planetary emergency (1). Climate concerns rightly prompted the 2015 Paris Agreement, which has facilitated coordinated global action not only among governments but also among companies, cities, and citizens. Research since then suggests that efforts to stabilize the climate and avoid the undesirable outcomes of >1.5°C warming will require a rapid reduction in land conversion and a moratorium by about 2035 (2). The most logical path to avoid the approaching crisis is maintaining and restoring at least 50% of the Earth’s land area as intact natural ecosystems, in combination with energy transition measures (2, 3). Those measures by themselves will likely be insufficient and must be augmented by restoration to create negative emissions to offset the likely clearing and release of greenhouse gases that will occur until a 2035 moratorium can be reached.

Natural ecosystems are key to maintaining human prosperity in a warming world (4, 5), and 65% of Paris Agreement signatories have committed to restoring or conserving ecosystems (6). Intact forests, and especially tropical forests, sequester twice as much carbon as planted monocultures (7, 8). These findings make forest conservation a critical approach to combat global warming. Because about two-thirds of all species on Earth are found in natural forests, maintaining intact forest is vital to prevent mass extinction (9). However, carbon sequestration and storage extends far beyond rainforests: Peatlands, tundra, mangroves, and ancient grasslands are also important carbon storehouses and conserve distinct assemblages of plants and animals. Further, the importance of intact habitats extends to the freshwater and marine realms, with studies pointing to least disturbed wetlands and coastal habitats being superior in their ability to store carbon when compared with more disturbed sites (10, 11).

Opportunities to address both climate change and the extinction crisis are time bound. Climate models show that we are approaching a tipping point: If current trends in habitat conversion and emissions do not peak by 2030, then it will become impossible to remain below 1.5°C (2, 12, 13). Similarly, if current land conversion rates, poaching of large animals, and other threats are not markedly slowed or halted in the next 10 years, “points of no return” will be reached for multiple ecosystems and species (13). It has become clear that beyond 1.5°C, the biology of the planet becomes gravely threatened because ecosystems literally begin to unravel (12, 14). Degradation of the natural environment also diminishes quality of life, threatens public health, and triggers human displacement because of lost access to clean drinking water, reduced irrigation of important subsistence crops, and exacerbation of climate-related storm and drought events (15). These occurrences will become increasingly worse without substantial action over the next few years. Additionally, human migrations, triggered by climate change–induced droughts and sea-level rise in combination with extreme weather events, could displace more than 100 million people by 2050, mostly in the southern hemisphere (12, 13).

A companion pact to the Paris Agreement—a Global Deal for Nature (GDN)—could help ensure that climate targets are met while preventing species extinctions and the rapid erosion of biodiversity and ecosystem services in the terrestrial, freshwater, and marine realms. The concept of a GDN as a policy mechanism emerged from an earlier study restricted to protecting biodiversity in the terrestrial realm (16). We expand that perspective to the freshwater and marine realms while simultaneously lending support to an alternative pathway to remaining below 1.5°C that relies heavily on aggressive conservation of remaining habitats. This approach not only safeguards biodiversity but also is the cheapest and fastest alternative for addressing climate change and is not beholden to developing carbon removal technologies unlikely to be effective or to scale in the time-bound nature of the current twin crises (4). Here, we offer a policy framework based on scientific guidelines that could pair nature and climate deals, be mutually reinforcing, and recommend time-bound milestones and targets. We identify specific threats and drivers of biodiversity loss, and discuss costs of implementation of a GDN. Finally, we introduce breakthrough technologies for monitoring progress.


Atmospheric sciences, Earth sciences, and remote sensing provide the main scientific and technological basis supporting the Paris Climate Agreement. The science of conservation biology underpins the GDN and is based on five fundamental goals: (1) represent all native ecosystem types and successional stages across their natural range of variation—or “representation”; (2) maintain viable populations of all native species in natural patterns of abundance and distribution—or “saving species”; (3) maintain ecological function and ecosystem services; (4) maximize carbon sequestration by natural ecosystems; and (5) address environmental change to maintain evolutionary processes and adapt to the impacts of climate change (17).

A practical application of many previous researchers has been to determine the spatial dimensions required to support goals 1, 2, 3, and 5 within large biogeographic units such as ecoregions—ecosystems of regional extent, terrestrial, freshwater, or marine—or regions holding concentrations of endemic species (biodiversity hotspots), the latter of which are typically delineated as clusters of ecoregions (16, 18). Carbon sequestration is a more recent concern.

Protected areas are the cornerstone of biodiversity conservation, and studies document that well-managed reserves are far more effective in safeguarding biodiversity than are other forms of land use (19). Studies across ecoregions and other large units show that to achieve the five goals of conservation biology requires in the range of 25 to 75% of land or water under some form of conservation management (2022). Several efforts have identified a middle ground of 50% protected to ensure conservation of biodiversity and ecosystem services upon which humanity depends (16, 20, 23, 24). Exactly how much should be protected to maintain the diversity of life and secure the benefits nature provides is an empirical question that is best determined per biogeographic unit—from the ground up—to better inform global numerical targets. This finer spatial scale delineation adds efficiency by accounting for ecoregion-level patterns of endemism, beta-diversity, and connectivity requirements to maintain viable populations of area-sensitive species (16).

To ensure representation of native ecosystem types, goal 1, terrestrial ecoregions have been a widely used ecological classification system for conservation planning for nearly three decades (25). The rationale is that a global map of ecoregions can serve as the framework to ensure creation of networks of protected areas that represent the widest array of habitats and, by extension, conserve the widest range of species and their unique adaptations to their environments. A recent global review tested the distribution of more than 200 million species records of plants, animals, and fungi against the map of terrestrial ecoregions (Fig. 1A) and revealed sharp, statistically significant discontinuities in species ranges across boundaries (25). Thus, ecoregions effectively represent similar clusters of not only habitat types but also species—underpinning analyses to address goals 1 and 2 of the GDN (25).

Fig. 1 The world’s 846 terrestrial ecoregions and depiction of 30% protection by the 2030 milestone.

(A) The 846 terrestrial ecoregions. (B) Levels of protection by 2030.

To ensure that global protection efforts reinforce conserving viable populations (goal 2), many studies over the past two decades provide digital distributional data on well-known vertebrate taxa, some plants, a few invertebrate groups, and their levels of endangerment [e.g. (26)]. Many of these studies mapped distributions of taxa with limited ranges (endemics) or highly threatened species. Studies vary in spatial scale from identifying species limited to a single site, to endemics confined to narrow ranges, or to larger aggregations of rare or threatened species, and have been amalgamated within frameworks such as Key Biodiversity Areas (KBAs) or larger Biodiversity Hotspots (18, 27). At the other extreme are wide-ranging species that occur at low densities and often require extensive areas complemented with use management to maintain viable populations. Wide-ranging or area-sensitive large mammalian herbivores and carnivores require stepping stone or contiguous habitat between reserves to allow migrations, seasonal movements, and gene flow (28). Connectivity among reserves also becomes vital for ensuring species persistence in a changing world and for fostering ecosystem resilience (goals 3 and 5) (29).

The principles described above apply to the ocean as well. Marine ecoregions, defined for both coastal and pelagic provinces, delimit regions with distinct species assemblages often characterized by regional endemics (30, 31). Marine protected areas (MPAs), and in particular fully protected marine reserves, have proven much more effective than other actions (e.g., fisheries management) in protecting and restoring biodiversity, increasing yields in adjacent fisheries, and enhancing ecosystem resilience (32). In addition, management of fisheries through either marine reserves or effort control produces identical yields under a reasonable set of assumptions (33), providing support for half of the ocean closed to fishing. On average, species richness is 21% higher and biomass of fish is six times greater within marine reserves than in adjacent unprotected areas (24). Marine reserves help restore the complexity of ecosystems through a chain of ecological effects (trophic cascades) once the abundance of large animals recovers sufficiently (34). Although marine reserves work best for species with limited movement ranges, large reserves can also protect large predators that conduct transoceanic migrations not only if reserves capture reproduction and nursery habitats but also through genetic selection (35). Marine reserves may not be immune to the effects of climate change, but reserves with complex ecosystems are more resilient than unprotected areas (36). Studies suggest that, on average, more than 30% of the ocean should be protected to achieve a series of environmental and socioeconomic objectives (22).

The current global protected area targets agreed by the Convention on Biological Diversity under what is called “Aichi Target 11” set coverage targets for the year 2020—17% in the terrestrial realm and 10% in the marine realm. These are interim measures that are politically driven but not science based and are widely viewed in the scientific literature as inadequate to avoid extinctions or halt the erosion of biodiversity (20). Compounding the problem is that only about half of the 14.9% of the terrestrial realm currently covered by protected areas is also connected (37). Significantly increasing the percent of connected protected area networks under a GDN will thus be essential to achieve representation while also ensuring that viable populations are conserved and ecological and evolutionary processes are maintained (37).

A growing body of research documenting the inherent interconnection between carbon sequestration and biodiversity lends further support for a proposal to pair a GDN with the Paris Accord. Carbon-rich ecosystems, by definition, sequester (both store and pump) the most carbon from the atmosphere. This carbon sequestration service is key to climate stabilization and to climate change mitigation. It is no coincidence that some of the most carbon-rich ecosystems on land—natural forests—also harbor high levels of biodiversity. Evolution has generated carbon-rich forests by packing in long-lived trees that also feed stable soil carbon storage pools. This packing effect is made possible by high levels of coexistence among diverse species and growth forms, and this coexistence has been made possible by the biotic interactions that generate competition and defense. It is the very pests, pathogens, pollinators, decomposers, and predators that comprise a tropical forest that generated the carbon-rich growth forms (in both wood and soil) that take the carbon out of the atmosphere.

Even in the world’s widespread savannas, carbon sequestration is enhanced by biodiversity. Herbivores are key to plant growth as well as soil carbon sequestration. Predators are key to maintaining herbivore balance with primary production. Too many herbivores result in lower carbon storage, but too few predators result in overinvestment in plant biomass, which leads to off-scale fires and losses to the atmosphere. In the ocean realm, from coral reef to blue water, biodiversity is part and parcel to the flux of atmospheric carbon to stored carbonates and deep ocean sediments. Without biodiversity, this system shuts down as well. Studies across multiple realms demonstrate a loss of carbon potential as the biodiverse systems are degraded or destroyed (5, 10, 11).

The nexus of climate and biodiversity science offers chilling scenarios of the unraveling of biotic systems if temperatures exceed a 1.5°C rise in average global temperature (14). Climate models and vegetation models often underestimate the sensitivity of ecosystems due to idiosyncratic relationships between species. Clear examples include (i) coral polyps rejecting their algal symbionts causing bleaching events, (ii) native bark beetles producing massive tree mortality in coniferous forests, and (iii) debilitating infestation of ticks affecting reindeer, moose, and other cold-climate large ungulates. Historically, we know that individual species were able to move during past climate swings such that ecosystems disassemble and surviving species assemble into novel configurations. However, in the current climate crisis and with reduced connectivity of natural landscapes, species may be unable to move fast enough to track shifting climatic envelopes or at all (38). Tropical rain forests will likely revert to savanna-like vegetation in some species-rich regions, and tropical cloud forests, home to a disproportionate number of endemic species, will be severely compromised by reduction in cloud-borne moisture. These strong feedback loops render the planet impossible to manage biologically. The time-bound targets of the GDN will have the greatest short-term effect on saving species and habitats deemed most sensitive to rapid climate change (39).


Our objective is to present scientific guidance for three major themes that should be included in a GDN and a short list of key milestones and targets that could underpin these themes, which would be complementary and, in many cases, reinforcing of the Paris Climate Agreement. These themes are (1) protecting biodiversity, (2) mitigating climate change, and (3) reducing threats to ecosystem intactness and persistence of species. We also propose that the GDN embrace monitoring progress from the ground, or below the sea surface, to space, using powerful new technologies, much of it publicly available.

Theme 1: Protecting biodiversity

We support calls to conserve at least 30% of the Earth’s surface by 2030 (40). This is viewed as a milestone toward the larger end goal of half of the planet protected by 2050, if not sooner, made elsewhere (16). The 30% by 2030 milestone has also been proposed by the International Union for the Conservation of Nature (IUCN) and its member organizations as a critical step for marine conservation (IUCN Resolution: WCC-2016-Res-050-EN).

These global milestones and targets are useful: They are easy to comprehend and help simplify policy and communications strategies. But because biodiversity is unevenly distributed, conservation biologists and planners must be careful to avoid two major risks inherent in a single global percentage: (i) adding more land to reach the global target that is similar to what is already well accounted for at the expense of underrepresented habitats and species, and (ii) the temptation by some governments to protect low-conflict areas that may be lower priority from a biodiversity perspective (41). Adhering to the first goal of biodiversity conservation—representation—greatly reduces these risks. This is the rationale for setting milestones and targets at both levels—global and ecoregional—and shows how the latter helps to operationalize the former to better conserve life on Earth (Table 1).

Table 1 Increased representation by biome of ecoregions achieving 30% protection by 2030.

View this table:

To illustrate how biodiversity might be protected under theme 1, we conducted two related assessments: (i) how using the 846 ecoregions as a “conservation template” can achieve increased representation of critical habitats and species in networks of well-connected protected areas, and (ii) how that 30% fraction could be distributed to better protect areas of concentrated terrestrial biodiversity—from narrow-range endemic species in fragmented landscapes to endangered large vertebrate populations ranging across intact primary forests and grasslands. The remaining 20% of intact and semi-intact land is also essential for climate targets to avoid negative impacts on humanity and species populations (15). In addition to expansion of protected areas, we propose climate stabilization areas (CSAs) that would meet the criteria for Other Effective Area-based Conservation Measures (OECMs), recently defined by the Parties to the Convention on Biological Diversity (42). Working in cohort with the Paris Agreement, CSAs would concentrate in habitats like mangroves, tundra, other peatlands, ancient grasslands, and boreal and tropical rainforest biomes that store vast reserves of carbon and other greenhouse gases, and prevent large-scale land cover change. These complementary approaches are detailed below.

To assess how reserve expansion to meet the 30% by 2030 milestone could improve representation, we sorted the world’s 846 terrestrial ecoregions into four categories (Fig. 1B). These categories were created by the intersection of the most recent database of protected areas (43), showing a total of 14.9% of the terrestrial realm with the amount of remaining habitat—totaling 49.9% of natural habitat across the 14 terrestrial biomes—and partitioned by ecoregion. This total sums all intact and semi-intact habitats but excludes permanent ice in Greenland and the highly degraded Sahara ecoregions (table S1). The categories are as follows:

1) Milestone Achieved: ≥30% protected (26% of 846 ecoregions; n = 219);

2) High Potential: less than 30% currently protected, but enough habitat remains outside protected areas to reach 30% by 2030 (41% of 846 ecoregions; n = 347);

3) Moderate Potential: the sum of area protected and habitat remaining outside protected areas is between 20 and 30%. These ecoregions could meet the 2030 target with some restoration (10% of 846 ecoregions; n = 88); and

4) Nature Imperiled: <20% total habitat remaining (protected + unprotected), often much less (23% of 846 ecoregions; n = 192).

Achieving ecoregion milestones of 30% protected by 2030 would greatly improve representation in the global targets (Tables 1 and 2). Of the 347 High Potential (category 2) ecoregions, a median of 47% habitat remains outside protected areas. This provides the possibility for changes in land designation to increase the total number of ecoregions achieving 30% protection from 219 (26% of total ecoregions) to 566 ecoregions (67% of the total). A further 88 ecoregions in category 3 (10% of total) can achieve the 30% by 2030 milestone with some restoration. The 566 ecoregions (categories 1 and 2) that have reached or could reach 30% protected by 2030 are distributed among all of the 14 terrestrial biomes. The most complete representation would fall within the Tundra, Boreal, and Tropical and Subtropical Coniferous biomes (Table 1). The least representation would be in the Tropical and Subtropical Dry Broadleaf Forests and the Temperate Grasslands, Savannas, and Shrublands biomes (Table 1), the latter of which is the center of global food production.

Table 2 Combined milestones and measurable targets for a GDN to better protect biodiversity and biosphere function in the terrestrial, freshwater, and marine realms.

View this table:

One hundred and thirteen ecoregions (13%) have already exceeded 50% protection. Yet, these rarely include the largest ecoregions containing the vast carbon reservoirs. Thus, a critical caveat in the representation approach is that some ecoregions may need much more area under protection to sustain species and processes and avoid biospheric feedback from release of greenhouse gases after conversion (see below discussion on CSAs for Amazon, Congo Basin, Southeast Asia, boreal, and tundra). At the other end of the spectrum, species extinctions will likely be most swift and severe in the 192 Nature Imperiled ecoregions, category 4 (Fig. 1B and table S1). Thus, efforts to bring the Nature Imperiled ecoregions to a protection level of 10% emerges as a clear 2030 restoration and recovery milestone. These ecoregions constitute known centers of endemism but have only a median of 4% protected habitat and 1% remaining outside protected areas. An approach to create a “species safety net”—to ensure the representation of vanishing biota ranging from single-site endemics to intact large mammal assemblages requiring large landscapes—is presented below (and more in greater detail in Fig. 2A).

Fig. 2 Increasing representation of important terrestrial, freshwater, and marine biodiversity sites for global 2030 targets.

(A) Terrestrial and freshwater biodiversity sites. (B) Marine biodiversity sites. RR, IUCN Sites of Range Rarity; TS, Threatened Species Sites.

Range-restricted and area-sensitive species and conservation of beta-diversity. Saving species and populations is a key feature of a GDN, targeting first the rarest, most range-limited species under the most threat for immediate conservation (44). To ensure that under the 30% by 2030 milestone these additions by ecoregion include irreplaceable sites (e.g., those with narrow endemic taxa found nowhere else), we intersected several widely used fine-scale maps of biodiversity distributions with the ecoregion map and the current database on terrestrial protected areas (45). For each of these overlays, we calculated the area for that theme found in each ecoregion that is currently unprotected but not accounted for by other overlays (to avoid double counting, as the spatial extent of these various mapping efforts depicting species rarity often overlaps considerably). We summed these extensions to present the percent and amount (in km2) added to the global total toward the 30% by 2030 milestone (Fig. 2). These data layers that are included inform milestones and targets (Table 2):

1) The Alliance for Zero Extinction (AZE) sites that are currently unprotected (46). AZE sites pinpoint extreme rarity by seeking to protect those birds, mammals, reptiles, amphibians, cycads, and conifers whose entire global distribution is limited to a single site and ranked as Critically Endangered or Endangered. The total is 928, of which 34% are unprotected (Table 2). All unprotected AZE sites, plus a 1-km buffer surrounding them, are included in the 2030 milestone. These 320 unprotected sites, plus buffer, add 0.46% to the expansion of the global terrestrial protected areas, or 619,490 km2 (Fig. 2A and fig. S1A). In the marine realm, unprotected AZE sites represent 9955 km2 and would contribute less than 0.1% to the expansion of the global MPAs (Fig. 2B).

2) A map of range rarity derived from the IUCN Red List data for the world’s birds, mammals, and amphibians, showing where sites supporting high densities of narrow-range species remain unprotected. The unprotected hotspots of range rarity add 0.12% to the expansion of the global terrestrial protected areas total percent or 162,491 km2, not including AZE sites (Fig. 2A and fig. S1B). In the ocean, conservation priority hotspots (top 5%) for range-restricted species (47) represent 14.7 million km2 and can account for 4.75% of the expansion of the global MPAs (Fig. 2B and fig. S1C).

3) A map of the density of threatened species from the IUCN Red List, showing where concentrations of species threatened with extinction remain unprotected. This layer includes nearly 64,000 species evaluated to date by the IUCN Red List, with almost 20,000 deemed as threatened with extinction. The unprotected polygons could add 0.03% to the expansion of the global terrestrial protected areas total percent or 34,631 km2 excluding overlapping data from the first two layers (Fig. 2A and fig. S1D). Conservation priority hotspots (top 5%) for threatened marine species (47) represent 8.5 million km2 and account for 2.8% of the expansion of the global MPAs (Fig. 2B and fig. S1E).

4) KBAs map globally important sites that meet a number of biodiversity criteria, including the presence of threatened or significant species (27). Though incomplete for some regions of the world (e.g. New Guinea), KBAs are identified by national constituencies using globally standardized criteria and quantitative thresholds and thus are useful to guide the expansion of protected area networks. Currently, approximately 15,000 such sites have been designated, with about 60% within IUCN category I to IV reserves (27). Adding in the unprotected portion of KBAs and a 1-km buffer expands the global terrestrial protected areas total by 1.3% percent or 1,759,341 km2 (Fig. 2A and fig. S1D). Unprotected KBAs in the oceans represent 5.8 million km2 and can contribute 1.6% to the expansion of the global MPAs (Fig. 2B and fig. S1G).

Summing the total area of all of these unprotected sites (plus 1-km buffers) from the first four layers, and removing double counting, adds 1.9% to the existing 14.9% of the terrestrial realm and 9.2% to the existing 7.3% of the marine realm already designated for protection (45). The unexpectedly small area contribution of these irreplaceable terrestrial sites is due to the high spatial overlap present among layers 1 to 4.

5) The remaining contribution to the milestone of 30% by 2030 is drawn from High Biodiversity Importance Ecoregions (HBIEs) that depict 455 of the most biologically distinct ecoregions by biome (fig. S1H). It is a slightly revised and updated map of the most biodiverse ecoregions on Earth (48). This is subdivided into three themes encompassing small fragments or larger landscapes that remain unprotected and not covered in the above four layers:

i) The 36 Biodiversity Hotspots are included in the HBIE layer and contain a combined total of 114 category 3 and category 4 ecoregions, the most threatened categories. Adding the remaining unprotected habitat fragments found in these ecoregions contributes 0.51% or 439,569 km2 to the expansion of the global terrestrial protected areas total (Fig. 2A and fig. S1I).

ii) Biodiversity Hotspots (fig. S1F) represent educated guesses (or sometimes comprehensive lists) by botanists of regions where plant endemism is concentrated globally. These lists are often based on counts of endemics at particular sites—a depiction known as alpha-diversity. Equally important as a conservation metric, but rarely measured, is beta-diversity—the replacement or turnover of the rosters of plant communities with distance or along elevational gradients (fig. S1J). High beta-diversity ecoregions are most common in five biomes: Tropical and Subtropical Moist Broadleaf Forests; Tropical and Subtropical Dry Broadleaf Forests; Mediterranean Climate Shrublands; Tropical and Subtropical Grasslands, Savannas, and Woodlands; and the tropical and subtropical ecoregions found in Desert and Xeric Shrublands. Many of the ecoregions in these five biomes overlap with the biodiversity hotspots; adding in unprotected habitat in some missing ecoregions adds 6.9% to the global total or 9,188,027 km2 (Fig. 2A).

iii) The strongholds of large wide-ranging species that require vast areas to maintain viable populations are another important metric of biodiversity value. HBIEs capture where these large mammals occur by overlaying a map of intact vertebrate assemblages (fig. S1K) (49). Adding in unprotected intact large mammal habitat in some missing ecoregions adds 8.5% to the global total or 11,304,493 km2 (Fig. 2A and fig. S1H).

Conserving the unprotected terrestrial areas selected above would add 17.7% to the existing 14.9% currently protected and exceed the 30% milestone while creating a more representative global system of protected areas that incorporate the major aspects of terrestrial biodiversity (Fig. 2A). Our approach uses what we consider to be the most important biodiversity features, but we recognize that there are multiple pathways to achieve a 30% protected milestone.

Complete protection requires closer to 50% of the Earth’s land surface achieved by combining the above analysis with 20% designated as CSAs (see below section and Table 2).

Wide-ranging megafauna and migration routes. The future survival and recovery of megafauna will, in particular, require additional strategies beyond those above for prioritizing sites that require immediate attention. Many of our Earth’s large terrestrial vertebrate species are suffering range collapse and extirpation of populations and could disappear by the end of this century or sooner if intensive habitat conversion, poaching, and overhunting continue given the slow rates of population recovery characteristic of most megafaunal species (50). Similar sharp declines are being observed in freshwater and marine large vertebrate species that figure prominently as protein sources for billions of people (51). Large mammal migrations, one of the most outstanding natural phenomena on Earth, are disrupted globally as traditional migratory routes are cut off and habitat lost (28). Ultimately, the loss of large herbivores, ecosystem engineers, and top carnivores can produce detrimental effects across entire food webs, termed “trophic cascades” (52).

GDN targets can address maintaining megafaunal populations and landscapes in multiple ways not only through direct protection but also by promoting connectivity (Table 2). The current CBD Aichi Target 11 calls for “well-connected systems of protected areas,” although only a rudimentary definition of “well-connected” and inadequate guidance on how to measure connectivity is provided (37).

Corridors and connectivity. A key target of the GDN would be to reconnect protected areas via corridors along environmental gradients, in riparian networks, and between megafaunal reserves (Table 2). The amount of area required and how this would affect the protection or restoration target should be subject to a major study based on the needs of the most wide-ranging, area-sensitive species. Replanting of native trees or simply allowing degraded forest lands to recover as forest corridors could create an ecological road map guiding where restoration can have the maximum benefits for biodiversity.

Old-growth habitats. Before the industrial revolution, primary or old-growth habitats covered most of Earth’s 846 terrestrial ecoregions. Today, these primary habitats, as represented by unlogged forests, ungrazed deserts, ancient grasslands, and savannas—and in the marine realm, the untrawled seafloor and unfished seamounts—are now remnants. These ancient repositories of rich and vulnerable biodiversity are optimal arenas for life-sustaining processes (53). Clear, time-bound milestones and targets for the above biodiversity features, including targets for old-growth forests and ever-wet forests, are included in the GDN and are drawn from the scientific literature supporting each target (Table 2). Protecting habitats that have low anthropogenic disturbance offers the most cost-effective approach to conserve the largest number of species and also for their climate resilience and should become obvious targets under the GDN (54). As a prime example, tropical forests occur on only 7% of the land area, yet they harbor more than half of the world’s known species and most of these are dependent on primary forest (9, 55). The buffering conditions of many complex “old-growth” habitats also enable local adaptation to climate change for many vulnerable species. These biologically important areas also serve as carbon repositories. For example, ancient grasslands are extremely species rich, including endemics (53), and they store approximately as much carbon globally as forests. Because most grassland carbon is stored belowground, it is a highly secure and reliable carbon sink, especially in the face of fire and other climate-sensitive disturbance factors (56).

The freshwater realm. Freshwater habitats harbor roughly one-third of all vertebrate species and 10% of global biodiversity, yet only cover 1% of the planet (57). Freshwater ecoregions particularly rich in species, endemics, and intact megafauna are priorities for biodiversity conservation, as are those where watersheds are largely intact and waterways flow freely (58). Freshwater ecosystems and biodiversity are among the most threatened on Earth (57). Many of the protection measures identified for terrestrial areas cover required actions for freshwater, and targets for the former benefit the latter. The GDN targets protecting and restoring 30% of the world’s freshwater ecoregions by 2030 as a vital milestone (Table 2).

The marine realm. A meta-analysis of 144 studies showed that, on average, 37% of the ocean should be strongly protected (preferably in fully protected marine reserves) to achieve a number of goals: protect biodiversity, ensure connectivity, avoid species collapse and adverse evolution, ensure sustainable fisheries, and benefit multiple stakeholders (22). More than 80% of these objectives can be met with 50% marine reserve protection (22). With only 2% of the world’s ocean currently in fully to highly protected marine reserves (59), these estimates highlight enormous gaps in marine conservation.

In addition to biodiversity, marine ecosystems are also important carbon sinks. Coastal ecosystem such as mangroves, salt marshes, and seagrass beds sequester astonishing amounts of carbon per hectare and thus play a disproportionately large role in the global capture and storage of carbon (60, 61). Although the spatial extent of the world’s 20 mangrove ecoregion complexes is minute compared to other biomes (300,294 km2 or 0.23%), conservation and restoration of mangroves are vital to climate and marine conservation scenarios, and buffer storm surges and hurricanes.

The ocean also harbors old-growth ecosystems equivalent to primary forests, such as unfished seamounts with deep corals thousands of years old (62). In the coastal zone, some seagrass species such as the Mediterranean Posidonia oceanica have complex belowground organs that help accrete sediment and bury carbon over millennia (63). Preserving the vegetative cover of these coastal habitats is important, but preserving their megafauna is essential, as studies suggest that intact predator populations are critical to maintaining or growing reserves of carbon stored in marine ecosystems (64). For example, tiger sharks in seagrass beds in Australia create a “landscape of fear,” where sea turtles and dugongs preferentially forage in seagrass microhabitats that are low in predation risk. The majority of the habitat, with high predation risk, has greater carbon stocks.

Ongoing efforts are identifying priority areas that will help protect biodiversity at multiple levels (from species to ecosystems), help produce food through fish spillover from fully protected reserves, and help mitigate climate change through the protection and restoration of important carbon sinks. Here, we present general targets and categories that should be prioritized as part of the GDN (Fig. 2B). The current shortfalls in ocean protection are presented in a revised map of the coastal ecoregions and pelagic provinces of the world, alongside the global system of MPAs (Fig. 3, A and B) (30, 31).

Fig. 3 Coastal ecoregions, pelagic provinces, and marine protected areas of the world oceans.

(A) Coastal ecoregions and pelagic provinces. (B) Map of marine protected areas.

Theme 2: Mitigating climate change

Conserving the carbon storehouses: CSAs. Protecting 30% of the Earth as high-priority conservation areas will be essential but insufficient for holding emissions below 1.5°C (2). Carbon storehouses, our terrestrial carbon sink, currently absorb one-quarter of emissions, and the conservation of natural habitats as stabilization areas under the GDN reinforces this role. Natural habitats outside protected areas can also be managed to maintain intactness and enhance conservation of area-sensitive species. These lands can be considered under the term OECMs (43). CSAs can fit under the OECM umbrella and contribute to area-based conservation targets. Simply put, CSAs are areas where conservation of vegetative cover occurs and greenhouse gas emissions are prevented, which can be achieved under various forms of land management. A key target of the GDN would enhance efforts to maintain at least 85% forest cover in critical areas such as the Amazon. Some parts of the planet work as a system and need to be managed for a high proportion of intactness to continue to function as a weather machine for the planet (65, 66).

These essential carbon storehouses could be retained and protected under CSAs. The tundra and boreal biomes, which together comprise 24,162,700 km2 (approximately 18%) of Earth’s terrestrial landmass, are exceptional because of their carbon storage [fig. S2A and (13)]. Examples of forests with high carbon density include the temperate rain forests of the Pacific Northwest United States and Canada, the temperate moist eucalypt forests in southeast Australia, Congo Basin peat swamp forests, and intact forest reserves in Malaysian Borneo [e.g., (67, 68)]. In recent years, more than 60% of global emissions from natural habitats (which accounted for 15% of annual greenhouse gas emissions), stemmed from clearing and fires in just two provinces—Riau province in Sumatra, Indonesia, and Matto Grosso state in Brazil. Protection for these and other “high-biomass forests” that are disproportionately important in climate mitigation could be enabled through their designation as CSAs.

Tundra and boreal biomes, aside from extensive belowground carbon storage, retain the largest areas of intact large mammal assemblages and are home to some of the most extensive and extant large mammal migrations on Earth (fig. S1K) (28, 49). Globally, the world’s intact large mammal assemblages coincide with the carbon storehouses and overlap substantially with the world’s remaining wilderness areas, excepting permanent ice in Greenland and in the heavily degraded, defaunated, and overgrazed ecoregions of the Sahara (fig. S2B) (15). However, preventing the transformation of large vegetated regions is not sufficient to maximize their carbon sequestration role. In some major ecosystems, the presence of large predators and herbivores helps store more carbon (69). For example, many large tropical trees with sizeable contributions to carbon stock rely on large vertebrates for seed dispersal and regeneration (70).

Conservation of tiger habitat provides another example of how CSAs could also help improve species protection and recovery efforts. Tigers are habitat generalists, but today, extant populations are largely restricted to forests, and many of these forested habitats lack formal protection under IUCN reserves. Approximately 434 protected areas fall with the tiger’s current range (fig. S3), not all of which contain breeding tigers, and not a single protected area is large enough to support a viable population over the long term (71). These adjacent reserves must be managed as metapopulations, populations linked by occasional dispersal, and so connectivity, and in many cases restoring forest corridors, is essential. The sum of all protected areas in the tiger range is 495,807 km2 or 42% of the total area of all 76 Tiger Conservation Landscapes combined (fig. S3). Between now and 2030, only a fraction of the remaining habitat outside formally protected areas will be incorporated into expanded or new reserves. Yet, enough habitat remains to achieve the target of a near-tripling the wild tiger population (72), and management of this habitat could be supported under the OECM umbrella as megafaunal landscapes in that part of the breeding tiger population, especially in India, that occurs outside formal protected areas. The payoff for climate stabilization is dramatic: An earlier study showed that forested areas that contain tigers have three times the carbon density compared to forests and degraded lands where tigers have been eradicated (71). Restoration in tiger habitat and other megafaunal landscapes could center on protecting remaining fragments of natural habitat, reconnecting and buffering them by restoring degraded lands, thus aligning with the Bonn Challenge that seeks to restore 350 million ha of forest by 2030 (73). The same rationale could be used to extend protection in high-carbon density habitats for gibbons and other primates, hornbills and other large fruit-eating birds, and fruit-eating bats.

The central role of indigenous lands. Potentially prominent among OECMs are indigenous peoples’ lands, which account for 37% of all remaining natural lands across the Earth, and these lands store >293 gigatons of carbon (74). Although many of these lands meet the definition of a protected area, many others may be appropriately characterized as OECMs. Here, the global policies articulated in the Paris Agreement and the proposed GDN merge with addressing human rights. The direction, insights, rights, and voices of indigenous peoples are essential but rarely published in scientific journals. The GDN could assist indigenous peoples, where requested, to keep lands intact—for hunting areas, protection of traditional lifestyles, or other features—and provide a mechanism to assist these communities with securing tenure rights. Supporting efforts to maintain these lands, many of which are critical to global terrestrial biodiversity conservation, in many cases would result in lower rates of deforestation and better protection of the biodiversity and ecosystem functions upon which these communities depend (75). Furthermore, more than one-third of the carbon identified in community lands across the tropics lies in areas without secure tenure rights. The Amazon Basin, Congo Basin, boreal, tundra, Borneo, and New Guinea ecoregions all store massive amounts of above- and belowground carbon (fig. S2A) and overlap greatly with indigenous lands.

Theme 3: Reducing major threats

Land conversion and infrastructure development risk compromising the ability of protected areas and CSAs to protect species and store carbon. Slowing and stopping the clearing of intact natural habitat for agriculture, the dominant form of land use today, is critical as part of the overall strategy to stay below 1.5°C. By increasing intensification and directing cropland expansion to degraded lands, and by reducing food waste, the 2050 world food demand could be met without additional land clearing (76, 77). The total length of paved roads globally is projected to increase by 20 million km (enough to encircle the Earth more than 600 times), and 90% of all new infrastructure is slated for the world’s tropical and subtropical biodiverse ecosystems (78). Infrastructure and energy development projects—major sources of fragmentation and penetration into wilderness areas, protected areas, indigenous territories, and CSAs—should be closely scrutinized (see Table 3 for recommended targets and policies) (79). Proactive approaches are needed to optimize human benefits while limiting harm.

Table 3 Enabling policies, milestones, and targets to reduce major threats and drivers of change.

View this table:

On land, hunting by humans imperils 40 to 50% of all threatened bird and mammal species (80). In the marine realm, industrial fishing is the largest hunting operation on the planet and targets more than half of the ocean surface, spanning an area four times that covered by terrestrial agriculture (81). Currently, fishing exploitation rates remain uncontrolled in vast ocean areas, including the high seas. Only a small fraction of the fisheries of the world are managed and science based, and they mostly concern single species targeted by industrial fleets in developed countries (82). In many cases, reducing fishing effort could help increase efficiency and profitability (83). Illegal and unsustainable trade in animals and plants, especially in threatened species, must also be curtailed. Further, resources will be needed to enforce protection as protected areas expand under a GDN. The most commonly invoked intervention to counteract poaching and overhunting is law enforcement patrolling to deter, detect, and punish poachers. Halting illegal and unsustainable trade in animals and plants, in particular of species threatened with extinction and where trade adds to the pressure on that species, is essential (Table 3).

The proliferation of invasive species, pollutants, and toxins is a major driver of species loss, population declines, and habitat degradation around the world (84). The amount of plastic making its way into the oceans is predicted to nearly double in the next decade; allowing this to occur would unleash extremely detrimental impacts on marine species and ecosystems (85). Beyond plastics, widespread use of ecologically damaging toxins is causing massive declines in global pollinators, invertebrate biomass, and degradation of aquatic ecosystems. To achieve the Sustainable Development Goals (SDG) target to prevent and significantly reduce pollution, the world needs to move from our current “linear economy” (make, use, dispose) to a circular economy in which resources do not become waste but instead are recovered and regenerated at the end of each service life (86). A GDN should encourage appropriate regulations, market incentives, and enterprise in areas such as waste management upstream to prevent plastic trash entering the ocean (Table 3). Funding for research, technology, and invasive species management programs in targeted areas can have marked effects in restoring native species populations and ecosystem services.

The conservation biology literature offers extensive analyses and detailed case studies of global threats and drivers of biodiversity loss. We have distilled key papers from this literature to identify clear milestones and targets that reflect their scope and intensity and how to reduce their impact as an integral part of the GDN (Table 3).


The Paris Agreement offers a useful template for a GDN because it sets global targets, provides a model for financial support, and supports bottom-up efforts. All nations have signed on to this agreement. But the Paris Agreement is only a half-deal; it will not alone save the diversity of life on Earth or conserve ecosystem services upon which humanity depends. It is also reliant on natural climate solutions that require bolstering outside of the Paris Agreement to ensure that these natural approaches can contribute to its success. Yet, land-based sequestration efforts receive only about 2.5% of climate mitigation dollars (4, 87).

At the same time that climate scientists were arriving at a single numerical target for maintaining Earth’s atmosphere at safe limits, biodiversity scientists identified multiple targets for the required habitats to conserve the rest of life on Earth. But to communicate effectively, as in the Paris Agreement, these many needs could be encompassed within a single target: protect at least half of Earth by 2050 and ensure that these areas are connected (16, 23, 40). The evidence arising since these calls were made clearly demonstrate that while we may be able to afford to wait to formally designate 50% protected in nature reserves, we need to fast-track the protection and restoration of all natural habitat by 2030 (2). A GDN that will ensure that we have at least 50% intact natural habitats by 2030 is the only path that will enable a climate-resilient future and is one that will offer a myriad of other benefits (3, 4). Since the crucial role of intact, diverse systems has also been demonstrated to be essential for carbon storage (8, 15), the GDN will need to emphasize mechanisms for protecting intactness both inside and outside of protected areas (e.g. in CSAs/OECMs) well before 2050.

Tallis and colleagues (3) demonstrated that with existing technologies and large-scale adoption of common conservation approaches (e.g., protected areas, renewable energy, sustainable fisheries management, and regenerative agriculture), it would be possible to advance a desired future of multiple economic and environmental objectives (including 50% of each biome intact, with the exception of temperate grasslands). This spatial coexistence is possible even with the prospects of feeding and supporting the material needs of a growing human population (88). The success of proposals to boost food production while protecting biodiversity will likely depend on our success in addressing human population growth, however, and our willingness to marshal financial resources accordingly (89).

Gross costs for nature conservation measures across half the Earth could be $100 billion per year, but the international community currently spends $4 billion to $10 billion per year on conservation (90). Extending the area-based targets in the post-2020 strategic plan for biodiversity to 30% by 2030 will likely require direct involvement of the private sector. In key sectors—fishing, forestry, agriculture, and insurance—corporations may be able to align their financial returns directly to reaching targets recommended by the GDN. However, the typical approach to conservation planning does not involve the real (net) costs because the direct benefits of conservation and the averted costs of inaction are not included in the calculations. Barbier and colleagues (90) showed that potential direct benefits from biodiversity conservation for various sectors range from increasing annual profits by $53 billion in the seafood industry to $4300 billion in the insurance industry. In addition, marine reserves can provide more economic benefits from tourism than fishing in many locations worldwide (91). Financial investments of even 10 to 20% of potential benefits from biodiversity conservation from three key industries could make up as much of one-third of the commitment needed to implement a GDN. A GDN may appeal to a broader set of nonstate actors, including corporations and local government entities. The solutions could be implemented in ways that have direct positive benefits to local or regional communities and especially indigenous peoples. Land-based jobs, food security, green space, access to wilderness, and ecosystem services are benefits that deliver advantages to rural and urban dwellers alike.

Complex life has existed on Earth for about 550 million years, and it is now threatened with the sixth mass extinction. If we fail to change course, it will take millions of years for Earth to recover an equivalent spectrum of biodiversity. Future generations of people will live in a biologically impoverished world. Adopting a GDN and the milestones and targets presented here would better allow humanity to develop a vibrant, low-impact economy and conserve intact ecosystems, all while leaving space for nature. Linking the GDN and the Paris Agreement could solve the two major challenges facing the biosphere and all the species within it and result in a return to safe operating space for humanity.


Supplementary material for this article is available at

Section S1. Maps of important biodiversity and carbon layers

Section S2. Underlying data for increasing representation of ecoregions by adding unprotected areas of high priority

Section S3. Monitoring progress under the GDN from the ground to below the sea to space

Fig. S1. Maps used to increase representation among terrestrial ecoregions and unprotected sites of biodiversity importance that contribute to the global milestone of 30% protected by 2030.

Fig. S2. Maps showing total terrestrial carbon and overlap with intact large mammal assemblages.

Fig. S3. Overlay of tiger conservation landscapes and protected areas.

Table S1. Underlying data for increasing representation of ecoregions by adding unprotected areas of high priority.

References (105109)

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited.


Acknowledgments: We thank B. Babbitt, M. Gomera, N. Lapham, S. Pimm, and J. Watson for reviewing drafts of the manuscript. Two anonymous reviewers helped improve the article. Author contributions: E.D., C.V., E.S., T.E.L., D.O., K.B., R.F.N., J.E.M.B., and Y.P.Z. designed the framework of this paper; A.R.J., S.F., J.M., and A.B. conducted data analysis and interpretation of data; E.D., C.V., E.S., D.O., G.P.A., N.D.B., R.F.N., S.F., K.B., T.B., N.H., L.N.J., and E.W. contributed substantially to the writing and editing of the manuscript. Competing interests: There are no financial or competing interests involved in authorship of this paper. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper, the Supplementary Materials, and/or the materials cited herein. Additional data related to this paper may be requested from the authors.

Stay Connected to Science Advances

Navigate This Article