Area-based conservation in the twenty-first century

Humanity will soon define a new era for nature—one that seeks to transform decades of underwhelming responses to the global biodiversity crisis. Area-based conservation efforts, which include both protected areas and other effective area-based conservation measures, are likely to extend and diversify. However, persistent shortfalls in ecological representation and management effectiveness diminish the potential role of area-based conservation in stemming biodiversity loss. Here we show how the expansion of protected areas by national governments since 2010 has had limited success in increasing the coverage across different elements of biodiversity (ecoregions, 12,056 threatened species, ‘Key Biodiversity Areas’ and wilderness areas) and ecosystem services (productive fisheries, and carbon services on land and sea). To be more successful after 2020, area-based conservation must contribute more effectively to meeting global biodiversity goals—ranging from preventing extinctions to retaining the most-intact ecosystems—and must better collaborate with the many Indigenous peoples, community groups and private initiatives that are central to the successful conservation of biodiversity. The long-term success of area-based conservation requires parties to the Convention on Biological Diversity to secure adequate financing, plan for climate change and make biodiversity conservation a far stronger part of land, water and sea management policies. The long-term success of area-based conservation—including both protected areas and other effective area-based conservation measures—after 2020 will depend on governments securing adequate funding and prioritizing biodiversity in land, water and sea management.


Review
concern that nations are paying less attention to the qualitative elements of Aichi Target 11, including the need for representative and connected protected areas that are equitably governed and managed [10][11][12][13][14][15][16] . Moreover, some long-standing issues-including poor resourcing and low management effectiveness-continue to compromise the ability of protected areas to conserve biodiversity and ecosystems [17][18][19][20][21][22][23] . As a consequence, there is a risk that humanity could fail to deliver on the overall strategic goal for which the target was established: to 'improve the status of biodiversity by safeguarding ecosystems, species, and genetic diversity' 6 .
A post-2020 global biodiversity framework will be agreed upon at the fifteenth conference of the parties to the CBD. This new strategic plan could be humanity's last chance to prevent catastrophic loss of global biodiversity 24 . The urgency to act has emboldened calls for a substantial expansion of area-based conservation globally [25][26][27][28] and fundamental changes in how environmental targets are framed and implemented 17,22,29,30 . It is therefore timely to assess the achievements and failures of area-based conservation efforts over the past decade, and to place these findings within the wider context of the global biodiversity crisis.

The performance of protected areas since 2010
Here we provide an up-to-date temporal analysis (spanning between 2010 and 2019) of how the recent expansion of protected areas globally has affected the net coverage of the qualitative components of Aichi Target 11; details of methodology and calculations are provided in the Supplementary Methods. We omit reference to OECMs in this Box 1

Protected areas and OECMs are complementary area-based conservation measures
A protected area is formally defined as "A clearly defined geographical space, recognised, dedicated and managed, through legal or other effective means, to achieve the long term conservation of nature with associated ecosystem services and cultural values" 2 ; an OECM is defined as "A geographically defined area other than a Protected Area, which is governed and managed in ways that achieve positive and sustained long-term outcomes for the in situ conservation of biodiversity, with associated ecosystem functions and services and, where applicable, cultural, spiritual, socioeconomic, and other locally relevant values" 89 . The feature that distinguishes between a protected area and an OECM is that the former has a primary conservation objective whereas the latter delivers the effective in situ conservation of biodiversity, regardless of its objectives. Protected areas help to conserve critically endangered species such as the Philippine eagle (Pithecophaga jefferyi) (photograph credit: B. Demeulemeester) (a) and the Nassau grouper (Epinephelus striatus) (photograph credit: E. Poliakoff) (b). OECMs have been recognized at a locally managed marine area on Totoya Island (Fiji) (photograph credit: S. Jupiter) (c) and at a conservation concession in Loreto region (Peru) (photograph credit: Brian Moriarty/CC-BY) (d).

Protected areas being ecologically representative
The concept of being ecologically representative has been interpreted as the coverage of species or ecoregions (areas that contain geographically distinct assemblages of species 31,32 ), especially those that are threatened with extinction 10,13,33,34 . We analysed how expansion of the global protected-area estate between 2010 and 2019 affected coverage of 12,056 species listed as 'Vulnerable', 'Endangered' or 'Critically Endangered' (hereafter referred to as threatened species) on the International Union for Conservation of Nature (IUCN) Red List 35 (Fig. 3). Between 2010 and 2019, the percentage of species with some portion of their geographical range protected increased from 86% to 87.6% (n = 10,563). However, only 21.7% (n = 2,618) of species assessed had adequate representation inside protected areas in 2019 (up from 18.9% in 2010), where adequacy targets for individual species were set according to their geographical range 34 .
The proportion of threatened reef-forming corals with adequate representation grew rapidly over the past decade, from 9.1% to 44.0%. The proportion of species with adequate coverage also increased for threatened mangroves (to 50.0%), seagrasses (to 50.0%), marine mammals (to 43.2%), marine bony fishes (to 42.1%) and cartilaginous fishes (to 32.4%) over this period. However, no threatened marine reptiles had adequate levels of protection in 2019. On land, the proportion of species with adequate coverage grew by <3% for birds (to 33.6%) and <2% for amphibians (to 10.9%), reptiles (to 13.6%), mammals (to 37.0%) and freshwater species (to 19.0%) in the past decade (Supplementary  Tables 1, 2). It remains that 78.3% (n = 9,438) of all threatened species assessed had inadequate protection as of 2019, with at least 1,493 (12.4%) remaining without any coverage at all.
We further assessed progress towards the adequacy targets of 17% protection for terrestrial ecoregions and 10% for marine ecoregions or pelagic regions. We found that 42.6% (n = 361) of terrestrial ecoregions were adequately protected in 2019 (up from 38.8% in 2010) (Fig. 3, Supplementary Table 3). We also found that one-fifth (20.6%) of land protected since 2010 covered tropical and subtropical grassland ecoregions-a critically endangered biome 31 (Supplementary Table 4). However, 32.9% of land protected since 2010 covered dry or desert ecoregions, which are relatively species-poor and are well-represented in the global protected-area estate 31 . The percentage of marine ecoregions with adequate coverage increased to 45.7% (n = 106) over the past decade (from 31.8% in 2010) (Supplementary Table 5 Table 6). The expansion of protected areas in pelagic regions was particularly concentrated in waters between Australia and South America, which received 36.6% of all new marine protected area (6.0 million km 2 ) in the past decade.

Coverage of areas of particular importance for biodiversity
The 'Key Biodiversity Area' (KBA) approach 36 offers a global standard for identifying marine, terrestrial and freshwater sites that contribute substantially to the global persistence of biodiversity. Over 15,000 KBAs have so far been identified (83.1% of which are 'Important Bird Areas', the avian subset of KBAs) 37 . Host nations are encouraged to ensure that these sites are managed in ways that ensure the persistence of biodiversity, although this does not necessarily mean inclusion within a protected area 36 . Our analysis showed average coverage of terrestrial KBAs was 45.9% in 2019 (up from 43.6% in 2010) and 43.3% for marine KBAs (up from 37.9% in 2010) (Fig. 3). Overall, around 4,900 KBAs (33.0%) remained without protected area coverage in 2019.
Wilderness areas are ecologically intact land and seascapes that are predominantly free of human-driven biophysical disturbance 38,39 . They underpin planetary life-support systems 40 and are critical for the long-term persistence of imperilled species 41 , especially in a time of climate change 42 . Over half (55.6%) of all wilderness overlaps with the geographical range of at least one threatened species, yet wilderness areas are also very spatially discordant from KBAs-only 1.2% of all land and sea on Earth is simultaneously recognized as both a KBA and a wilderness area ( Supplementary Fig. 1). Our analysis shows that coverage increased for both terrestrial (from 19.7% to 22.1%) and marine wilderness (2.0% to 8.5%) areas during the past decade (Fig. 3).

Coverage of ecosystem services
The carbon sequestered and stored in terrestrial ecosystems has a pivotal role in mitigating anthropogenic climate change 43 . We therefore The map shows the annual expansion of protected areas across marine (blue-to-pink colours) and terrestrial (green-to-red colours) realms on Earth. Protected area data were sourced from ref. 7 Fig. 2). However, maps of terrestrial organic carbon-particularly in peatlands and tropical rainforests-are continually being refined 45 , which may influence future understandings of carbon storage in particular areas.
The 'biological pump'-carbon fixed by phytoplankton in the oceans being exported to the deep ocean 46,47 -also has a key role in mitigating climate change, because it removes carbon from the ocean and atmosphere systems for decades to millennia 48 . We estimate that 0.21 Pg of particulate organic carbon 49 and 0.17 Pg of dissolved organic carbon 50 is exported inside marine protected areas each year (Supplementary  Tables 7, 8). However, the factors that drive carbon export in the oceans vary seasonally 51 and the relative value of marine protected areas in carbon export may vary through time.
About three billion people rely on wild-caught or farmed seafood as their primary source of protein, which makes the sustained provision of seafood a globally important goal 5 . We compared protected area coverage of the most-and least-productive marine regions for fisheries catch in the oceans, finding that coverage of the least-productive exclusive economic zones (those within the bottom 20% for annual fisheries catch per unit area) (Supplementary Table 9) was-on averagethree times greater than coverage of the most-productive exclusive economic zones (those within the top 20% for annual fisheries catch Ph ilip pi ne s S ou th K or ea In do ne si a P a k is ta n G e o rg ia Ir a n M y a n m a r V ie tn a m K y rg y z s ta n In d ia A z e r b a ij a n S a u d i A r a b ia B a n g la d e s h J o r d a n K a z a k h s t a n U z b e k is t a n T u r k m e n i s t a n O m a n N o r t h K o r e a I r a q Y e m e n S y r ia T u r k e y A f g h a n is t a n per unit area) (Supplementary Table 10) (31.2% versus 9.8%). Moreover, average coverage of the most-productive exclusive economic zones has not changed notably since 2010 (Fig. 3). We also found that the seven most-productive pelagic regions for fisheries catch have no formal protected area coverage (Supplementary Table 11).

Protected areas being well-connected
Well-connected ecosystems are critical for maintaining important ecological and evolutionary processes (including species migration and gene flow), especially when species face rapid climatic and environmental changes 52,53 . Connectivity among marine protected areas further helps to replenish and maintain fish populations, including on fished reefs 54,55 . A previous study showed that in 2016 only 30% of terrestrial ecoregions were at least 17% covered by protected areas that were within the potential dispersal distance of terrestrial vertebrates 56 .
A subsequent study showed that the percentage of connected terrestrial protected areas increased from 6.5% to 7.7% between 2010 and 2018 57 . However, these assessments did not account for the permeability of unprotected land between protected areas. There have been no global-scale assessments of connectivity among marine or freshwater protected areas, but regional-scale studies show them to have limited connectivity-especially for species with a dispersive larval stage 58 .

Effectiveness of protected area management
Over the past few decades, four broad approaches have been used to evaluate area-based conservation efforts ( Table 1). Three of these

Review
approaches pertain to management effectiveness, the first of which (which we term input evaluation) evaluates the adequacy of management resources for area-based conservation. A recent study of about 23% of terrestrial protected area found that 47% of protected areas suffer from inadequate staff and budget resources, with poor resourcing especially noticeable in the Neotropics 21 . Similarly, a study of 433 marine protected areas showed 65% to have insufficient budget for basic management needs and 91% to have on-site staff capacity that is inadequate or below optimum 20 . Related inputs, including the weak enforcement of protected area regulations 59 , have also been implicated in poor management effectiveness. A second evaluation approach (which we term threat reduction evaluation) asks whether area-based conservation effectively reduces threats to the persistence of biodiversity. The majority of these evaluations show that protected areas slow, but do not completely halt, human pressures within their borders. For example, human pressures increased inside 55% of protected areas on land between 1993 and 2009 60 , and in the marine realm 94% of protected areas created before 2014 permit fishing activities 61,62 . However, terrestrial protected areas have been found to reduce rates of deforestation and forest degradation to below those observed in nearby unprotected areas 63 , including in the Amazon 64,65 ; marine protected areas can reduce fishing vessel traffic 66 and the negative effects of some non-native species 67 .
The third evaluation approach (which we term outcome evaluation) asks whether the goals of area-based conservation are being achieved relative to no intervention taking place. A recent controlled study showed that around 12,000 protected areas were ineffective at reducing human pressures inside their borders between 1995 and 2010 68 . However, several studies have reported beneficial effects of protected areas on biodiversity. For example, a controlled study of 359 terrestrial protected areas showed species richness to be 10.6% higher and abundance 14.5% higher inside protected areas than outside, with the effects of protection being most prominent in human-dominated land uses in the tropics (for example, cropland and plantations) 69 . Similarly, a controlled study of 218 marine protected areas found that, on average, fish biomass inside protected areas is nearly double that in non-protected sites 20 . Marine protected areas can also promote the recovery of commercial fish species 70,71 . In particular, no-take marine reserves can effectively increase species richness, density and biomass in both tropical and temperate systems 20,72-74 , as well as being effective at restoring trophic function 75,76 and lowering levels of coral disease 77 . Finally, several studies have reported on the social effects of protected areas. For example, a controlled study covering 603 protected areas found that households near protected areas with tourism opportunities had higher wealth levels (by 17%) and a lower likelihood of poverty (by 16%) than similar households living far from protected areas 78 .

Equitable governance and management in protected areas
Social equity in the context of protected areas has multiple dimensions, including distributional equity (for example, people agree on a scheme for sharing benefits and burdens), procedural equity (for example, decision-making that is transparent, accountable and participatory) and recognition (for example, respect for cultural identities, customary rights and traditional management practices) 79 . A recent survey covering 225 protected areas showed the majority of conservation managers, staff and community representatives believe that the benefits of protected areas are shared equally 15 . Yet, the study also showed that decision-making was not equitable in many cases and that local stakeholders perceived a general loss of rights over natural resources after the establishment of a protected area 15 .
Despite limited evidence of progress towards social equity, protected areas that do integrate local communities as stakeholders often produce better socio-economic and conservation outcomes 14,80 . A review of 27 marine protected areas found that stakeholder engagement, surveillance, leadership, political will and the existence of sanctioning and conflict resolution mechanisms were key factors in achieving ecological objectives 81 . No-take, well-enforced and longer-established marine protected areas show not only conservation success 82 , but also positive economic and governance outcomes for dependant human communities 83 . Furthermore, community-managed terrestrial protected areas are often more effective than nationally designated protected areas at reducing deforestation pressures, including in Peru, Brazil, Australia and Namibia 84,85 .

Lessons learned and priority actions
Collectively, national governments have made some progress towards Aichi Target 11 in the past decade-particularly in the marine realm. However, it is clear that nations have, as yet, failed to meet this target. The rate of expansion of terrestrial protected areas needed to be double what was observed in the past decade to achieve 17% coverage for land The different approaches imply different measurements and are subject to strengths and weaknesses. Design, input and threat reduction evaluations all measure means to an end, whereas outcome evaluation measures progress towards the ultimate goals of area-based conservation. Globally, we have limited capacity to perform outcome evaluation for area-based conservation.
and freshwater environments. Moreover, 78.3% of known threatened species and more than half of all ecosystems on land and sea remained without adequate protection in 2019. A clear lesson from this assessment is that nations must expand area-based conservation efforts and better ensure that they contribute meaningfully to global goals for species and ecosystem conservation, which range from stopping extinction 86 to keeping ecosystems intact 87 . The past decade has also shown that many protected areas are poorly managed (owing predominantly to chronic resource shortages), and that many Indigenous and community groups are inadequately or inequitably represented in land, water and sea conservation plans. In light of these lessons, we identify three urgent challenges that must be acted upon by governments, scientists, policy-makers and other stakeholders as they embark on the next decade of area-based conservation (Tables 2, 3).

Making OECMs count
There are now expanding opportunities to formally recognize places outside state-run protected areas that can conserve biodiversity. In addition to protected areas that are governed privately 88 and by Indigenous peoples 84 , other OECMs are being increasingly recognized. The importance of OECMs was formally recognized in Aichi Target 11 in 2010, but their guiding principles and criteria for identification were not agreed until November 2018 (Box 1). This delay probably contributed to OECMs being overlooked in most national biodiversity policies and strategies over the past decade. With a formal definition now agreed 89 , nations and managing bodies look set to operationalize OECMs more rapidly. The challenge now for the conservation community is to ensure that OECMs contribute meaningfully to biodiversity conservation.
OECMs could help to address representation shortfalls in the global protected-area estate. One recent study has shown that 566 unprotected KBAs are at least partly covered by one or more potential OECMs 4 , andcompared with nationally designated protected areas-OECMs may prove to be more socially acceptable in productive land-and seascapes (which are hotspots for poorly protected threatened species 10,11 ). Recognizing OECMs in inshore marine habitats, farmlands and managed forests could also enhance the connectivity of area-based conservation efforts, providing that natural ecological functions can be restored and maintained in such areas 90,91 . Wider recognition of OECMs should also help to make area-based conservation management more equitable, as they are managed by and for the benefit of a diverse set of actors. A recent study showed that lands managed by Indigenous communities in Australia, Brazil and Canada support concentrations of vertebrate species similar to those inside nationally designated protected areas 92 , which exemplifies the importance of working with Indigenous peoples to recognize OECMs in their territories.
However, to deliver on the potential of OECMs, governments, private industry and the conservation community must immediately mobilize support for OECMs to overcome the issues that are faced by many protected areas, including inadequate reporting and resourcing. A reporting platform for OECMs 8 was released in December 2019 and has the potential to make assessments of progress towards the successor of Aichi Target 11 more accurate, if countries make use of it. The success of OECMs will also depend on governments and conservation actors upholding human rights and social safeguards, particularly in Indigenous and community areas. In cases in which meeting OECM criteria will require some adaptation to livelihoods, great care must be taken to develop alternative livelihood opportunities that deliver tangible benefits to resource users 93 . Alternative livelihood schemes must also be mindful to retain the biodiversity benefits of OECMs 94 .

Tracking the increasing dynamism of area-based conservation
Recent studies have shown that protected areas are more dynamic in space and time than previously thought 9 . Decisions to remove, shrink or relax protected areas are poorly documented, which makes it difficult to assess which ecosystems are most susceptible to such dynamics or how these changes affect the overall quality of area-based conservation networks. The challenge for the conservation community is to have protected area dynamics reported more transparently, especially when they compromise biodiversity outcomes.
Many removals from the protected-area estate can be attributed to 'protected area downgrading, downsizing and degazettement' (PADDD) events. More than 1,500 PADDD events affected, in total, over one-third of the protected area network in Australia (416,740 km 2 ) between 1997 and 2014 95 . Moreover, 23 PADDD events have affected natural World Heritage sites (protected areas with 'outstanding universal value', such as the Virunga, Serengeti and Yosemite National Parks) 96 . PADDD events can accelerate forest loss and fragmentation 97 , and most (62%) are associated with activities that are in stark conflict with biodiversity conservation, including industrial-scale resource extraction and infrastructure development 98 . Potentially of greatest concern are the many PADDD events that are going undocumented, particularly in marine systems 99 and on private lands 100 .
To improve the transparency of area-based conservation decisions, we encourage governments and the conservation community to engage more with global PADDD tracking platforms (for example, www.padddtracker.org). We also believe that integrating PADDD tracking data with existing area-based conservation databases (for example, the World Database on Protected Areas 7 ) would vastly improve their utility and aid global reporting. Dynamism in area-based conservation could signal attempts to expand or enhance protected areas, either through improved resourcing and management 101,102 or by enacting more restrictive regulations 103 . As such, there is also a clear need to better incentivize and track the continuum of changes to protected areas that can improve their ability to conserve biodiversity. We suggest that such changes be characterized collectively as 'protected area gazettement, expansion To conserve: • 17% of terrestrial and inland water X • 10% of coastal and marine areas X To capture important places for biodiversity and ecosystem services, such as: • KBAs X • Wilderness areas X To be effectively managed by: • Having adequate resources X • Abating human pressures X • Having positive biodiversity outcomes X To be equitably managed X To be ecologically representative by: • Covering 17% of all terrestrial ecoregions X • Covering 10% of all marine ecoregions X • Covering 10% of all pelagic regions X To be well-connected and integrated X Progress towards targets is assessed as good (substantial positive trends at a global scale relating to most aspects of the element); moderate (the overall global trend is positive but insubstantial or insufficient; there may be substantial positive trends for some aspects of the element but little or no progress for others; or the trends are positive in some geographical regions but not in others); poor (little or no progress towards the element or movement away from it; although there may be local, national or case-specific successes and positive trends for some aspects, the overall global trend shows little or negative progress); or unknown (insufficient information to score progress). Review and enhancement' (PAGEE) events. Clear, transparent tracking around both PADDD and PAGEE events will ensure we address-and do not exacerbate-current shortfalls in area-based conservation.

Outcome-orientated evaluation of area-based conservation
The numerous approaches that have been developed to evaluate area-based conservation efforts all have merit, but the conservation community remains too reliant on types of evaluation that focus on management inputs or threat reduction 104 (Table 1). Adopting evaluation techniques that more effectively capture the biodiversity and socio-economic outcomes of area-based conservation is currently a substantial challenge. The Global Database on Protected Area Management Effectiveness provides information on many protected area processes, including the existence of a management plan or the adequacy of law enforcement activities 105 . Yet the majority of the methodologies of the Global Database on Protected Area Management Effectiveness were not developed to quantify the effects of protected area management activities on species and ecosystems 105 , and therefore cannot be used to evaluate progress towards the effective conservation of biodiversity. High-resolution maps of ecological change across land-and seascapes, including forest cover change 106 and changes in cumulative human pressure 107,108 , enable more outcome-orientated conservation evaluations. But ecological changes across land-and seascapes do not always explain local biodiversity patterns 109 . The temporal resolution of cumulative human-pressure mapping also lags behind that of forest-cover mapping efforts, and some maps of human pressure are at spatial resolutions (for example, 77 km 2 (ref. 108 )) that preclude assessments of many small (<1 km 2 )-but crucially important 110 -protected areas.
To make area-based conservation evaluations more outcomeorientated, we suggest making robust outcome evaluation techniques-both ex ante to help to site areas 111 and ex post to report on outcomes 112 -a standard reporting requirement for all organizations that are involved in area-based conservation. Conservation agencies must then better fund long-term and well-designed biodiversity monitoring programmes. It will be especially important for such programmes to monitor control sites that match protected areas in terms of ecological and anthropogenic conditions, so as to isolate any confounding effects 113,114 . We also encourage governments and communities to engage more with citizen science initiatives that use techniques such as camera traps, drones and acoustic monitors to increase the coverage and frequency of biodiversity data 115 . Combining advances in remote sensing and field campaigns (for example, the Biomass Earth Explorer of the European Space Agency) with in situ reporting of protected area management capacity and biodiversity trends (for example, using the Spatial Monitoring and Reporting Tool (SMART) could also make outcome evaluations much more accurate and reliable.

Future-proofing area-based conservation
Although the three challenges already discussed are immediate priorities, broader policy changes can ensure that area-based conservation can contribute meaningfully to longer-term goals held by the CBD: namely, that 'by 2050, biodiversity is valued, conserved, restored and widely used, maintaining ecosystem services, sustaining a healthy planet and delivering benefits essential for all people' 116 (Tables 2, 3). In this final section, we outline a set of necessary pre-conditions-adequate financing, being climate-smart and mainstreaming biodiversity across national policy frameworks-that require action by governments now to ensure the long-term success of area-based conservation strategies.

Secure adequate financing
The global funding available for species protection has more than halved in the past two decades, from approximately US$200 million per year in the 2000s to less than US$100 million per year in the 2010s 117 . The costs associated with expanding area-based conservation efforts compound resource shortfalls at existing sites. One estimate suggests that protecting and effectively managing a terrestrial protected-area network that is more taxonomically comprehensive that the current network would cost US$76.1 billion annually 118 . As such, a conservative estimate of the current financial shortfall for area-based conservation probably exceeds the multi-billion dollar mark. This shortfall is unlikely to be fully addressed in the coming decade, but reducing it must become an immediate priority for governments and private industry.
Current and future resourcing needs could be met if the contribution of area-based conservation to national economies was adequately recognized. The direct value generated by visits to protected areas is valued at US$600 billion per year 119 . Governments must therefore better account for the contribution of area-based conservation efforts to national economies. When budgeting for area-based conservation, we suggest that governments use predictive measures of funding requirements and effects 120 , and that they consider the under-appreciated cost-saving benefits of effective biodiversity conservation. For example, it would be useful to compare the costs arising from the socio-economic devastation caused by zoonotic diseases such as severe acute respiratory syndrome or coronavirus disease 2019 with those needed to effectively manage area-based conservation networks in a way that reduces supply to illegal wildlife markets 121,122 . There is also an urgent need to better harness industry and philanthropic contributions to area-based conservation, through-for example-improved To secure adequate resourcing To be climate-smart To make biodiversity conservation mainstream Fund the contribution of area-based conservation to national economies Better harness industry and philanthropic contributions Safeguard ecological integrity Use decision-support tools to make robust decisions Adopt a bold overarching goal for biodiversity Adopt biodiversity accounting protocols that align compensation with desired trajectories for imperilled species or ecosystems 1 Overarching preconditions require action by governments to ensure the long-term success of area-based conservation strategies.
funding guidelines that ensure that the involvement of private interests does not compromise the siting or management of area-based conservation [123][124][125] .

Being climate-smart
Anthropogenic climate change will become an increasingly strong mediator of the success of area-based conservation this decade 126 , with many predicted biological responses to climate change already underway 127 . A recent study showed that under a business-as-usual scenario for greenhouse gas emissions (Representative Concentration Pathway (RCP) 8.5), mean sea-surface temperatures within marine protected areas are projected to increase by 2.8 °C by 2100 128 . We overlaid climate change projections under a more-moderate emissions scenario (RCP 4.5) on the terrestrial protected-area network and found that temperatures in the warmest quarter will increase on average by 2.9 °C on protected land by 2050, with higher increases occurring in European nations ( Supplementary Fig. 2, Supplementary Table 12). We also found that by 2050 some biodiverse nations, including Suriname and Guyana, can expect 30-40% less rainfall on protected land during dry months ( Supplementary Fig. 3, Supplementary Table 13). Such changes in bioclimatic conditions are likely to markedly alter ecological networks 129 and imperil some species and ecosystems 130 . Even the relatively moderate RCP 4.5 scenarios are likely to drive the elimination of most warm-water coral reefs by 2040-2050 131 .
There are now well-established ways to incorporate climate change into area-based conservation plans 132 , including safeguarding or-where possible-restoring the integrity of ecosystems around protected areas so as to ensure ecological connectivity 133,134 . Ensuring that managers have the knowledge and capacity to implement realistic climate adaptation policies is also paramount 135 . Greater enforcement of conservation regulations 136 and accounting for human responses to climate change 137 are also likely to enhance the climate resilience of area-based conservation efforts. However, targeting protected areas in sites in which bioclimatic changes may be small (refugia) must be done with caution, because predicted changes can be spatially discordant 128 as can the biotic response. For example, only 3.5% of marine protected areas co-occur with refugia for both sea-surface temperature and oxygen concentration 128 . In such cases, decision-support tools (for example, value-of-information analysis 138 and systems modelling 139 ) can evaluate the benefits of resolving uncertainty about ecological responses to climate change before implementing conservation action, and therefore lead to more-robust management decisions.

Make biodiversity conservation mainstream
No matter how well-sited, resourced or managed, area-based conservation can act only on a subset of threats to biodiversity persistence. The amelioration of large-scale distal threats requires other interventions that are triggered by broader land, water and sea management policies 140,141 . China is the first major economy to formulate a national policy-known as the Ecological Redline Policy-that mandates municipality and provincial governments to establish biodiversity and ecosystem service assessments in land-use planning 142 . It is hoped that the Ecological Redline Policy extends to China's planned activities beyond their national boundaries-including the Belt and Road Initiative, which could affect many areas of critical conservation concern 143 . However, most national land, water and sea management policies are subservient to economic development 144 or contain loopholes that lead to perverse environmental outcomes 145 . Governments must recognize that getting these policies right is essential and will ease the strain on area-based conservation strategies in the long term.
Two cross-cutting changes could improve the efficacy of national land, water and sea management policies. First, we suggest that nations adopt an overarching goal for biodiversity that is bold-to have a net positive effect on biodiversity, for example 29 -and then agree a set of socio-economic and environmental targets that can contribute proportionally to this overarching goal. Targets should then be made mutually conditional: environmental targets (for example, to protect 30% of land) cannot be considered met if progress towards socio-economic targets (for example, to eliminate incentives harmful to biodiversity) is found to be wanting. Improving biodiversity accounting protocols could also enhance the efficacy of land, water and sea management policies. One example of this is switching from biodiversity impact-offsetting protocols that simply displace conservation funding or entrench rates of biodiversity loss (for example, averted loss offsetting) to emerging protocols that align compensation with desired trajectories for imperilled species or ecosystems (for example, target-based compensation) 146 .

Conclusions
Area-based conservation will remain the cornerstone of biodiversity conservation long into the twenty-first century, but governments have markedly underinvested in protected areas and OECMs and been weak in legally protecting them. In addition to addressing existing shortfalls, conservation organizations need to adopt more impact-orientated evaluation measures and promote governance and management equity. Organizations must also improve the transparency of decisions that result in spatial and resource dynamics, and ensure that OECMs can contribute meaningfully to biodiversity conservation. Finally, governments must future-proof area-based conservation by securing adequate financing, being climate-smart and mainstreaming biodiversity across environmental and socio-economic policies.

Code availability
Spatial analysis was conducted using ESRI ArcGIS Pro v.2.4.0. The workflow we used to process the World Database on Protected Areas is available in the Supplementary Information.