Maximizing the effectiveness of national commitments to protected area expansion for conserving biodiversity and ecosystem carbon under climate change

Abstract Global commitments to protected area expansion should prioritize opportunities to protect climate refugia and ecosystems which store high levels of irrecoverable carbon, as key components of an effective response to biodiversity loss and climate change. The United States and Canada are responsible for one‐sixth of global greenhouse gas emissions but hold extensive natural ecosystems that store globally significant above‐ and below‐ground carbon. Canada has initiated a process of protected area network expansion in concert with efforts at reconciliation with Indigenous Peoples, and acknowledged nature‐based solutions as a key aspect of climate change mitigation. The US, although not a party to global biodiversity conventions, has recently committed to protecting 30% of its extent by 2030 and achieving the UNFCCC Paris Agreement's mitigation targets. The opportunities afforded by these dual biodiversity conservation and climate commitments require coordinated national and regional policies to ensure that new protected areas maximize biodiversity‐focused adaptation and nature‐based mitigation opportunities. We address how global commitments can best inform national policy initiatives which build on existing agency mandates for regional planning and species conservation. Previous analyses of global conservation priorities under climate change have been tenuously linked to policy contexts of individual nations and have lacked information on refugia due to limitations of globally available datasets. Comparison and synthesis of predictions from a range of recently developed refugia metrics allow such data to inform planning despite substantial uncertainty arising from contrasting model assumptions and inputs. A case study for endangered species planning for old‐forest‐associated species in the US Pacific Northwest demonstrates how regional planning can be nested hierarchically within national biodiversity‐focused adaptation and nature‐based mitigation strategies which integrate refugia, connectivity, and ecosystem carbon metrics to holistically evaluate the role of different land designations and where carbon mitigation and protection of biodiversity's resilience to climate change can be aligned.


| INTRODUC TI ON
The accelerating pace of biodiversity loss and climate change have prompted increasing calls for transformative change, including expansion of the global protected area network (Grumbine & Xu, 2021;IPBES, 2019). Although substantial progress has been made toward increasing protected area coverage globally, new areas have often not been sited with regard for their efficacy in safeguarding biodiversity or resilience to climate change (Díaz et al., 2019). Leaders of both the United States and Canada have joined those of other nations in endorsing global calls to protect at least 30% of their respective nations (ECCC, 2020a;White House, 2021). However, translating such commitments into national and subnational policies that maximize the role of new protected areas in addressing both climate change and biodiversity loss is challenging (Arneth et al., 2020).
Here we explore how expansion of the protected area network within Canada and the United States can more effectively address climate-change-related threats to biodiversity by protecting climate refugia and areas with high levels of carbon stored within intact (i.e., high ecological integrity) ecosystems.
Protected areas are an essential component of a broader suite of strategies for sustainable coexistence between humans and the natural world, which includes restoration of degraded lands and conservation management of agricultural landscapes (Figure 1) . Protected area networks designed without information on how climate-change-related threats vary across the landscape may poorly protect biodiversity into the future (Carroll & Noss, 2020;Carroll et al., 2017;Stralberg, Arseneault, et al., 2020).
Priorities based on current patterns of biodiversity may fail to capture important locations if, for example, species shifts toward cooler climates increase the conservation value of high elevation areas Holsinger et al., 2019;Rowland et al., 2016).
Certain areas play a disproportionate role in allowing landscapes to retain their capacity to support native species and ecosystems in the face of climate change (the focus of climate adaptation planning in the biodiversity context used here). Climate refugia (areas buffered from climatic shifts where organisms can persist permanently or temporarily) represent "slow lanes for climate change" which temporarily limits biodiversity loss while nations work to reduce their greenhouse gas emissions . Refugia are created by a variety of factors ranging from local topography to broad-scale climate gradients Morelli et al., 2020;Stralberg et al., 2018). Climate connectivity areas or corridors also play a key role by facilitating dispersal to newly climatically suitable habitat, which allows such habitat to augment the role of refugia Parks et al., 2020). Although climate refugia are not the only landscape element important to biodiversity conservation, they are an essential conservation feature which is increasingly feasible to K E Y W O R D S climate change adaptation, climate change mitigation, climate velocity, connectivity, conservation planning, protected areas, refugia | 3397 CARROLL And RAY identify from broad-scale spatial data Stralberg, Carroll, et al., 2020).
Increased attention is also needed to the essential role of natural ecosystems in ameliorating climate change through their capacity to remove carbon dioxide from the atmosphere and offset direct emissions from other sectors (Goldstein et al., 2020;Morecroft et al., 2019). Proactive protection of above-and belowground carbon within intact ecosystems provides an important mitigation pathway distinct from that offered by improved intensive management or restoration of agricultural, grassland, and forested landscapes, such as widely publicized tree planting initiatives (Fargione et al., 2018;Griscom et al., 2017;Law et al., 2018;Seddon et al., 2021). An approach that focuses on avoided conversion of natural areas at meaningful scales is especially important in landscapes such as peatlands and old forests that hold large quantities of ecosystem carbon that, if lost due to disturbance, would be irrecoverable within a timescale meaningful to addressing climate change (Beaulne et al., 2021;Goldstein et al., 2020;Law et al., 2018). The large amount of land required, the convergence of biodiversity and climate change agendas (Roberts et al., 2020), and feedbacks between biodiversity loss and climate change (Arneth et al., 2020) should compel planners to seek to holistically align objectives for new protected areas that maximize co-benefits .
Previous analyses concerning global conservation priorities under climate change (Dinerstein et al., 2019(Dinerstein et al., , 2020Jung et al., 2020;Maxwell et al., 2020;Yang et al., 2020) have largely lacked information on refugia locations due to limitations of globally available datasets, and have been only tenuously linked to the policy contexts of individual nations and regions. In this essay, we provide recommendations that can inform commitments for protected area network expansion at various scales so that they encompass areas that contribute to achieving climate and biodiversity targets. We first describe how existing global conservation targets address climate adaptation and mitigation goals, and how these can in turn best inform national and subnational protected area strategies. We evaluate high-level policy opportunities for and barriers to achieving adaptation and mitigation goals at national and regional scales. We consider whether available data on refugia locations are adequate for guiding protected area establishment, and to what extent these locations overlap with priority areas for conserving ecosystem carbon. Finally, we use a case study to demonstrate how regional biodiversity-focused adaptation and nature-based mitigation strategies can be nested within national analyses of the role of different land ownerships and designations.

| THE G LOBAL CONTE X T: PROTEC TED ARE A AND MITI G ATI ON TARG E TS IN INTERNATI ONAL TRE ATIE S
The first global commitment to biodiversity conservation was enshrined in 2010-2020 targets developed by the Conference of the Parties (COP) to the Convention on Biological Diversity (CBD) (CBD, 2010). The CBD has been ratified by 193 UN member states with the notable exception of the US. Although the CBD is a legally binding multilateral instrument, in practice the CBD's goals and targets advance conservation outcomes at the national level by stimulating ambition via pressure from civil society organizations, the scientific community, and peer policymakers in other signatory nations. Even within a nation such as the US which is not a party to the CBD, global targets indirectly influence policy proposals generated by policymakers and NGOs (White House, 2021).
The CBD's overarching goal is to reduce the rate of biodiversity loss and safeguard nature's contributions to people in an equitable manner . The agreement to be finalized at the CBD's forthcoming meeting (COP15), termed the Global Biodiversity Framework (GBF), will include targets and indicators designed to comprehensively reflect the status of global biodiversity (CBD, 2021). The diversity of life is commonly recognized as being expressed at several scales: diversity among ecosystems, among species, and genetic diversity within species. Recent proposals have suggested that the GBF include targets for reducing the rate of loss and eventually halting loss of diversity at each of the three scales of biodiversity . For example, proposed targets related to biodiversity status and trend include no net loss of intact ecosystems, reduction in species extinction rates, and retention of 90% of existing genetic diversity within species (Figure 1) Laikre et al., 2020;Watson et al., 2020).
The most widely publicized biodiversity-related commitment adopted by the CBD involves area-based targets ( Figure 1). These include a 2010 commitment to protect 17% of terrestrial areas in ecologically representative and connected networks under strict protection or "other effective area-based conservation measures" (OECM) (CBD, 2010;Jonas et al., 2014). This target may be raised to 30% at the CBD's forthcoming Conference of the Parties (CBD, 2021). Approximately 12% of the terrestrial US and Canada is currently strictly protected, but no standardized estimates are available for the extent of OECM lands (ECCC, 2020d;USGS, 2018).
The selection of targets and indicators included within the GBF, although informed by science, is ultimately a political process with negotiated outcomes. The goals of the framework include both comprehensively tracking the status of biodiversity at all levels and communicating this status to a wide audience of policymakers and the public. There is no single index of the status of biodiversity akin to the IPCC's focus on the magnitude of global mean temperature shifts. Area-based targets such as 17% have been adopted by the CBD in part because they are easily communicated and measured, and have been linked conceptually and empirically to ecologically based targets such as reduction in species extinctions (Bhola et al., 2021;Noss et al., 2012). In fact, several recent publications have proposed that the 30% target is likely insufficient, particularly in geographies with more intensive human land use (Allan et al., 2019;Dinerstein et al., 2019;Noss et al., 2012).
Because biodiversity and the threats it faces are distributed unevenly across the landscape, the location of newly protected and conserved areas helps determine whether achieving an area-based target results in effective progress toward the overarching goal of halting biodiversity loss, and secondary goals such as reduced species extinction rates . The GBF specifies that monitoring of progress toward area-based targets should also evaluate the spatial importance of new protected areas, in terms of both site characteristics (e.g., does the protected area capture hotspots of species richness?) and emergent properties of the protected area system as a whole (e.g., is the resulting network representative and connected?) ( Figure 1).
Both area-based targets and spatial data on priority areas (e.g., refugia) can inform conservation actions such as protected area establishment ( Figure 1). With sufficient resources and effective management, newly protected areas will in theory contribute to improvements in status and trend indicators and achievement of associated targets (e.g., no net loss of ecosystems). However, due to the complexity of ecological systems, this connection may be indirect and subject to substantial time lags (Figure 1). Area-based targets and spatial data on conservation priorities are themselves based on incomplete information and should be revisited and iteratively refined based on information derived from monitoring of biodiversity status and trends ( Figure 1).
Global systems such as the Key Biodiversity Area (KBA) standard are intended in part to support site selection and make increases in protected area coverage more effective in improving biodiversity status and trends (Smith et al., 2019). Similarly, Dinerstein et al. (2019Dinerstein et al. ( , 2020) described a global network of priority areas that would fulfill the GBF's 30% area-based targets via protection of sites with high species diversity which as a whole would represent all ecoregions and form a well-connected network of sites. Nations Sustainable Development Goals contains both biodiversity and climate goals, and recognizes that protecting natural areas for adaptation and mitigation can also yield collateral ecosystem services by ensuring clean water supplies and buffering developed landscapes from extreme weather events such as flooding (Díaz et al., 2019Zeng et al., 2020). The draft post-2020 GBF also includes targets for such ecosystem services (CBD, 2021).
The GBF is novel in that it creates specific targets that provide opportunities to potentially align UNFCCC mitigation commitments and CBD biodiversity-related goals (Arneth et al., 2020). The GBF's Target 7 proposes that by 2030, "nature-based solutions and ecosystems-based approaches" will provide increased contributions to climate change mitigation and adaption (CBD, 2021). Achievement of this target would be monitored in part via trends in carbon stocks in different ecosystems, and the contribution of natural ecosystems to climate change adaptation. The goal of greater coordination of climate adaptation and mitigation actions in the context of biodiversity conservation has received support from the High Ambition Coalition, an alliance of ~50 countries (including Canada) within the UNFCCC, and from the 65 CBD member states that are signatories to the 2020 Pledge for Nature calling for a unified "One Health" approach to addressing biodiversity, climate, and environmental issues (UNEP, 2020).

| HOW C AN NATIONAL COMMITMENTS TO PROTEC TED ARE A E XPAN S I ON EFFEC TIVELY ADDRE SS B IODIVER S IT Y LOSS AND CLIMATE CHANG E?
Policymakers encounter several challenges when they attempt to incorporate climate adaptation and mitigation priorities into protected area expansion efforts. Firstly, limited guidance is available on how global targets and indicators can best stimulate and inform national and subnational policy and regional conservation plans. Secondly, our knowledge on how to identify areas important for biodiversity, let alone climate refugia and corridors, and hence track progress toward their combined protection, is imperfect and constantly evolving. Lastly, priority areas for adaptation and mitigation may occur in different regions or require different management regimes, raising questions about how best to coordinate the two goals with one another and with other considerations related to biodiversity and ecosystem services.
Because area-based targets such as 30% are simple and relatively easy to track, they have received most attention in national commitments (Díaz et al., 2019;Maxwell et al., 2020;Visconti et al., 2019). However, an exclusive focus on area-based targets can lead to a mismatch between the location of new protected areas and regions of high value to biodiversity, adaptation, and mitigation (Carrasco et al., 2021;Maxwell et al., 2020;Morecroft et al., 2019;Visconti et al., 2019). Because many factors are considered in siting protected areas, recent studies have found that new conservation areas established in response to global commitments have had limited success in capturing areas of highest importance to current biodiversity  or climate refugia (Carrasco et al., 2021). Conversely, unlike in the case of area-based targets, tracking progress toward adaptation and mitigation goals (e.g., COP15 Target 7) is difficult in part because of a scarcity of well-developed metrics for measuring achievement.
Although previous publications concerning global conservation priorities under climate change provide essential context, they have been limited by globally available datasets (Dinerstein et al., 2019(Dinerstein et al., , 2020Jung et al., 2020;Maxwell et al., 2020;Yang et al., 2020). The databases used in conservation planning processes (including global prioritizations) typically include information on current distribution of globally threatened taxa as proxies for biodiversity; they rarely include spatial data on climate refugia or climate connectivity areas, or address the additional uncertainty that arises from projecting species' response to climate change.
Priority setting at national or subnational extents can take advantage of a wider range of data on climate refugia and other adaptation targets than is available at the global extent ( Figure 2). Nonetheless such data are more limited than information on current biodiversity patterns, in part because they are based on recently developed metrics (Table 1). Whereas the ecological responses of species to climate change are complex, the metrics used to map climate refugia are based on simplified representations of these dynamics. In this essay, we compare several major categories of metrics commonly used to identify refugia, climate corridors, and ecosystem carbon (Table 1, Figure 2; see Text S1 for definition of metrics).

F I G U R E 2
Range of patterns shown by nine metrics relevant to biodiversity-focused climate adaptation and nature-based mitigation in North America, categorized by type of input data and underlying model. All data were rescaled to equal-area quantiles ranging from low (blue) to high (red) for comparability. See Supporting Information S1 for data sources and definitions of metrics Colour figure can be viewed at wileyonlinelibrary.com] Although not all of the referenced publications specifically mention refugia, the methods listed can be used to identify refugia as areas where the current and projected future distribution of a feature overlap or are in proximity. See Supporting Information for further description of metrics

| CHALLENGES IN IDENTIFYING ADAPTATION PRIORITY AREAS AND TRACKING PROGRESS TOWARD THEIR PROTECTION
An ideal framework for tracking and protecting biodiversity under climate change would include targets and indicators which are coherently linked across scales of biodiversity. The GBF's three-level framework (ecosystems, species, and genes) can be complemented by another categorization of ecological indicators, the three-track framework for systematic conservation planning (Noss et al., 2002;Noss & Cooperrider, 1994). This latter approach suggests that protected area networks be designed to adequately represent both coarse-filter or non-species-specific landscape elements (e.g., vegetation types) and fine-filter or species-specific habitat features, and also fulfill requirements (such as connectivity) necessary for persistence of key ecosystem processes and the population viability of focal species such as large carnivores. The three-track framework encompasses indicators of biodiversity composition (the coarse-and fine-filter elements correspond to composition at the ecosystem and species scales, respectively), as well as function (focal species viability). As Díaz et al. (2020) (p. S16) stated, "Ecosystem integrity, currently defined to include functional, compositional, and structural/spatial components, is more elusive to monitor than ecosystem area, but no less crucial for the long-term continuity of ecosystem functioning".
We can integrate these two indicator frameworks to provide a comprehensive categorization of the varied approaches to identifying climate refugia (Table 1). At each of the three scales of biodiversity, practitioners can integrate information from metrics which do or do not incorporate species-specific data, as well as metrics based on more complex process-based or mechanistic models if available.
Since all of the metrics shown in Table 1 have their strengths and weaknesses, comparison of spatial priorities suggested by multiple metrics is informative.
Practitioners identifying refugia and other priority areas at the broadest scale of biodiversity, that of ecosystems and communities, can apply information from each of these three categories (Table 1). The most common approach at the level of biomes uses process-based models such as dynamic global vegetation models (DGVM), which simulate shifts in potential vegetation in response to climate Gonzalez et al., 2010). Finer-scale spatial units in this global ecosystem classification hierarchy, such as ecoregions, have been identified based on biogeographic and environmental discontinuities, and can be effective tools for representing current patterns of biodiversity at a global scale . Because process-based models such as DVGM have not been applied to simulate shifts in such finer-scale ecological types, their relative resilience and vulnerability to climate change is assessed using statistical models which identify a climatic niche for each ecoregion and project the niche forward under future climates   (Table 1). Although such ecoregion models do not directly consider species-specific data, researchers have recently developed a metric termed the Bioclimatic Ecosystem Resilience Index that relates data on spatial turnover in species composition to existing and projected future environmental conditions (Ferrier et al., 2020).
At the next scale of biodiversity, that of species, a wide variety of approaches is available to identify priority areas under climate change. The draft GBF proposes that results from species-specific climatic niche models for a large number of taxa be aggregated and used to create global maps of the relative vulnerability and resilience of biodiversity under climate change . In place of such statistical niche models, a variety of more complex mechanistic models of species range shifts have been developed. Some incorporate simple dispersal kernels while others are full spatially explicit population models (Heinrichs et al., 2019;Miller & McGill, 2018;Phillips et al., 2008).
Recent reviews have proposed that the GBF also include a focus on intraspecific biodiversity, for example, via a commitment to conserve 90% of existing genetic diversity Laikre et al., 2020). For species for which there are sufficient data, genetic viability under changing climates can be projected using a mechanistic simulation model such as a spatially explicit population model (Pierson et al., 2015). Alternately, statistical models associating distinct genotypes with specific environments can be projected to future climates to identify genotype-specific refugia . Where genetic data are lacking, climatic subdivisions of a species' range can serve as surrogates for intraspecific ecotypes . Planners can work to preserve adaptive capacity under climate change by conserving populations in all significant subunits of a species environmental niche .
Coarse-filter or non-species-specific approaches to identifying refugia for locally adapted populations often make use of climate velocity metrics Hamann et al., 2015). Multivariate climate space is divided into thousands of unique types, and distance between each climate type's current and projected future locations is measured. Refugia where a climate type's current and projected future locations overlap or are in close proximity allow locally adapted populations to remain within their suitable climate tolerances as climate changes . Various forms of climate-velocity-based metrics (e.g., outbound vs. inbound; Kling et al. (2020)) identify different types of coarse-filter refugia (Supporting Information S1).
All the approaches described above use future climate projections, and are consequently limited by the spatial resolution of those projections to identifying macrorefugia (areas where broad-scale climate is relatively stable and suitable for persistence). Microrefugia (small areas with locally favorable environments within otherwise unsuitable climates) may be important to persistence of species with modest area requirements under climate change (Dobrowski, 2011).
Because microrefugia are often created by terrain-related factors, topographic diversity (topodiversity) data are useful for identifying areas where a heterogeneous physical environment (e.g., steep elevation gradients or diverse aspects) increases the likelihood that species will be able to find nearby suitable habitat as climate changes .

| EMERG ENT CHAR AC TERIS TI C S OF THE PROTEC TED ARE A NE T WORK A S A WHOLE
Because the warming effects of carbon emissions are global, locating priority areas for protection in those ecosystems which hold and sequester the most carbon maximizes protection of the global land sink (Goldstein et al., 2020). This goal may contrast with biodiversity conservation strategies which, in addition to locating particular areas important for supporting requirements of individual species, require distributing such areas widely in order to maximize biodiversity representation within protected areas (Margules & Pressey, 2000;Watson et al., 2011). The GBF monitoring framework includes the goal that protected area networks be representative, as measured by indicators based on the distribution of ecosystems and species (Faith et al., 2008).
Recent reviews have proposed stratifying or tracking achievement of global area-based targets by ecoregion (Dinerstein et al., 2019) or by biome, a higher-level aggregation of globally defined ecoregions  An additional potential concern is how actions at smaller extents will be coordinated to achieve broader goals such as facilitating connectivity. The goal of increasing connectivity between existing protected areas must be balanced with efforts to protect ecosystems in regions which are currently underrepresented in the protected areas network. Novel elements of connectivity also must be considered, such as the climate corridors which facilitate species' range shifts under climate change . Climate corridors are areas that form the best route between current and future locations of specific climate types. Because dispersing organisms may need to avoid hostile climates, these routes are often circuitous rather than the straight-line paths (Dobrowski & Parks, 2016).
Models that predict climate refugia and corridors based on future climate projections (inner circle, Figure 2) are inherently uncertain due to variation between global climate models and policy uncertainty regarding future rates of greenhouse gas emissions (Belote et al., 2018). (Metrics shown here were calculated as anomalies from the current (1981-2010) projected temperature and precipitation to the 2071-2100 period, based on an ensemble mean of 15 representative CMIP5 AOGCMs for representative concentration pathway (RCP) 8.5; Wang et al., 2016). Such uncertainty in input data is compounded by that inherent in the assumptions of the climate velocity (Ordonez & Williams, 2013) and connectivity models themselves (outer circle, Figure 2). Metrics based on physical habitat data such as elevation ( Figure 2) avoid uncertainty related to projecting future climate, but retain substantial uncertainty regarding their accuracy at representing the complex factors which generate topographically driven microrefugia Dobrowski, 2011;Lawler et al., 2015).
Climatic niche models, which predict refugia are based on correlations between species distributions and current climatic conditions, provide a useful approximation of potential shifts in species' distribution in response to climate change (Wiens et al., 2009). However, they assume that current species distributions are at equilibrium in respect to climate, and ignore that the effect of climate change on many species is mediated by sympatric species and ecosystem structural components. For example, in the US Pacific Northwest region discussed in the case study below, the effect of climate change on many species is initially limited by the ecological inertia and microclimates created by large old trees Carroll et al., 2010;Perry et al., 2011). Subsequent shifts may be abrupt as fire disturbance regimes amplified by climate trigger ecosystem transitions . Conservation of fire refugia (areas disturbed less frequently or less severely by wildfire due to topography or other factors) in such systems is an important element of climate adaptation strategies Meddens et al., 2018).
These types of model uncertainty are not unique to climate adaptation planning, being also encountered by conservation planners when they attempt to optimize placement of new protected areas for multiple aspects of biodiversity despite lacking information on the distribution and ecological requirements of a large proportion of species (Reside et al., 2018;Watson et al., 2011). Such uncertainty can be addressed via quantitative methods that down-weight priority areas with uncertain conservation value (Kujala et al., 2013;Moilanen et al., 2006) or via guidelines for factoring uncertainty into site-level management strategies (Belote et al., 2017). Comparative evaluation of priorities identified by different metrics in the context of the known strengths and limitations of the underlying models (Figures 4-7) is often more informative than attempting to identify a single optimal metric, and incentivizes holistic management strategies which address all facets of biodiversity's resilience to climate change (Carroll & Noss, 2020;Díaz et al., 2020).
A major component of the GBF has involved development of metrics to track achievement of global biodiversity targets (CBD, 2021). However, many of the metrics we discuss here, such as climatic niche models, are challenging to develop as consistent global datasets at a resolution relevant to national or regional planning, due to computational challenges or limited input data. Rather than propose new adaptation-related GBF indicators that could be used to track progress globally, or evaluating the sufficiency of particular area-based targets, we focus here on demonstrating how national and regional datasets can be used to maximize the effectiveness with which area-based global targets (e.g., 30%) advance adaptation and mitigation goals. To supplement ongoing work at the global level, national-level efforts should include an inventory of relevant spatial data on refugia and connectivity, and guidance for its use by agencies in regional planning. Formal or informal learning networks, such as the Landscape Conservation Cooperatives described below, can play a key role in comparing and learning from processes in different regions.

| CHALLENG E S TO COORDINATING ADAP TATION AND MITIG ATION
Some recent proposals envision that the 30% of the landscape devoted to biodiversity protection could be distinct from "climate stabilization areas" managed for conservation of ecosystem carbon (Dinerstein et al., 2019(Dinerstein et al., , 2020. However, given the extent of land considered in these proposals, it is important to co-locate new protected areas based on both adaptation and mitigation goals where possible, while integrating these in a manner consistent with wider biodiversity and sustainable development goals (Morecroft et al., 2019). There are frequently commonalities between optimal management strategies for conservation of climate refugia and irrecoverable carbon as both benefit from maintaining intact ecosystems. For example, old trees in coastal forests of the US Pacific Northwest and other regions store substantial carbon while providing refugia for rare and endemic species in part by moderating local microclimate (Spies et al., 2019). In boreal North America (e.g., the Hudson Bay Lowlands), protecting carbon-rich boreal peatlands also safeguards key climate refugia due to the moderating microclimatic influence of mesic areas (Stralberg, Arseneault, et al., 2020).
In reality, commonalities between ecosystem carbon and either current biodiversity or refugia are often weak, especially in arid environments (Di . At the extent of the United States and Canada as a whole, we found that climate adaptation priority areas identified by Stralberg, Carroll, et al. (2020) are effectively uncorrelated with both above-ground carbon  and soil carbon (Hengl et al., 2017) (rank correlation on 0.016 and −0.031, respectively). We acknowledge that the metrics used here to measure carbon stocks do not fully represent the diversity of metrics that can be used to assess the carbon mitigation potential from conservation investments (e.g., Goldstein et al., 2020). Additionally, there may be tradeoffs between management for adaptation and mitigation, for example, in those ecosystems where suppression of fire disturbance provides short-term increases in carbon retention but negatively affects native species (Perry et al., 2011). The complexities of carbon management in such ecosystems will require planners to consider a range of management strategies, including more intensive management than tends to be associated with strictly protected areas (Belote et al., 2017).
Additionality and permanence are core aspects of quality assurance of climate change mitigation processes, particularly carbon offset projects (Federici et al., 2017). The imperative to consider both creates a potential mismatch between conservation of irrecoverable carbon within intact ecosystems and market-based systems for tracking mitigation commitments. To receive payments for naturebased mitigation via conservation of ecosystem carbon, managers must typically demonstrate that actions (e.g., protected area designation) result in avoidance of otherwise likely additional carbon loss (Federici et al., 2017). Pre-existing biodiversity conservation commitments within newly protected areas can hamper the ability to demonstrate additionality of carbon conservation agreements.
Protection of old forest from timber harvest in regions with infrequent natural disturbance provides measurable additionality (Buotte et al., 2020). In contrast, it is often difficult to establish additionality resulting from protection of below-ground carbon (e.g., boreal peatland complexes) unless the area is under immediate threat from mining or other ground-disturbing activities.
Due to the stochastic nature of large disturbance events, it is difficult to predict the risk of loss of above-ground carbon to fire in many forest ecosystems (Anderegg et al., 2020). This makes it challenging to demonstrate permanence of claimed carbon benefits for

F I G U R E 3
Overlap between the 30% of area with highest value for (1) climate adaptation, as identified by Stralberg, Carroll, et al. (2020) from a composite prioritization based on 5 climate refugia and connectivity metrics, and (2)  any single protected area. Carbon markets currently account for unexpected carbon loss from fire by tapping a reserve of offset credits.
The National Greenhouse Gas Inventory (NGHGI) framework developed under the Paris Agreement may offer a more appropriate method for tracking carbon benefits from protected area expansion while accounting for short-term fluctuations due to disturbance (Federici et al., 2017).

| PROTEC TED ARE A E XPAN S I ON THROUG H RECON CILIATI ON -BA S ED L AND US E PL ANNING IN C ANADA
Although the CBD's area-based targets have helped propel a global expansion of protected area networks, the pace of expansion has varied widely among nations . Canada and the United States, as an adopter and non-adopter, respectively, of global biodiversity conventions, provide an illustrative contrast. The two nations are responsible for one-sixth of global greenhouse gas emissions but also hold extensive areas where natural ecosystems store globally significant above-and below-ground carbon (Coristine et al., 2019). The US is among the 17 "megadiverse" countries of the world (Pariona, 2018) and Canada contains the second highest remaining area of intact ecosystems globally (Coristine et al., 2019;Watson et al., 2018  Nations that did not sign the UFA, other recent court decisions ruled that they still have title and rights and that these cannot be negatively affected without consent. The alignment of the federal Target 1 process with regional land use planning provides a major opportunity for   We illustrate how multiple refugia-related metrics can be holistically assessed in planning processes using visual aids such as star- multiple metrics into a single diagram to produce a composite "fingerprint" representing the varying magnitudes of factors affecting climate adaptation and mitigation values (Figure 4b) (Carroll & Noss, 2020;Garcia et al., 2014). Starplots for the Yukon (Figure 4), which represent the average conservation value of a metric in each planning unit, scaled in comparison to the range of values shown across the entire region (e.g., the Yukon territory), provide an example of how the climate adaptation and mitigation metrics reviewed here can inform planning.
The starplot patterns (Figure 4b) suggest that ecosystems with highest soil carbon lie within the taiga and tundra regions in the northern Yukon, whereas areas with greatest value as climate refugia are found in areas of higher topographic relief in the central and southwest Yukon (Figure 4c). Areas important for climate connectivity (in the non-species-specific form considered here) are found throughout the region. Although the continental-extent data we present here will be too coarse-resolution for some planning processes, several efforts are completed or ongoing to develop climateadaptation-related data at higher spatial resolution to help inform conservation planning co-led by First Nations in the Yukon (e.g., BEACONS, 2017;Cooke, 2017;Stralberg, Arseneault, et al., 2020).
Canada's effort to reach protected area targets promises to achieve ing carbon-related payments to help support local communities in lieu of revenue from extractive industries such as logging, oil and mining (Townsend et al., 2020). These mechanisms may also be applicable to other rural economies dependent on revenue from public lands.
While opportunities for Indigenous-led conservation and conservation of public (Crown) land are significant in Canada, the fragmented nature of jurisdiction and the primacy of natural resource development for revenue create significant barriers to progress (Ray et al., 2021). Nevertheless, the globally significant carbon storehouses (including the second largest peatland complex on the planet, the Hudson Bay Lowlands) and sizeable areas characterized by high ecological integrity (Grantham et al., 2020) offer pressing reasons to actualize government commitments.

| PROTEC TI ON OF REFUG IA AND C ARBON -RI CH ECOSYS TEMS VIA E XECUTIVE B R AN CH AC TI ON IN THE U N ITE D S TATE S
The incoming US federal administration has committed to the target of 30% protection by 2030 (White House, 2021). Policymakers have called for focusing protection on lands which would aid in preventing extinction, stabilizing ecosystems, and sequestering carbon and greenhouse gas emissions. This "30x30" initiative offers an opportunity to advance both adaptation and mitigation goals as the US simultaneously recommits to the UNFCCC's Paris Agreement (Rosa & Malcom, 2020). The confirmation of the first Indigenous Secretary of the US Department of the Interior has placed increased focus on how tribal co-management approaches can jointly address reconciliation and climate issues, and such approaches have been highlighted in recent executive orders (White House, 2021).
The separation of powers between the US executive and legislative branches frequently results in a divided government that presents barriers to ratification of international agreements and legislation related to climate mitigation and adaptation (Snape, 2009). In anticipation of such hurdles, the Paris Agreement was structured so that ratification by the legislative branch would not be required for US en- Comprehensive landscape planning for climate adaptation and mitigation must overcome barriers to coordination between multiple federal agencies with contrasting land management mandates (Mihm, 2000). A second requirement is sufficient political will. For example, pressure for expansion of renewable energy infrastructure on federal lands as part of federal efforts to reduce carbon emissions prompted development of the Desert Renewable Energy Conservation Plan (DRECP) to coordinate biodiversity conservation in in California's southeastern desert with expansion of solar, wind, and geothermal infrastructure (Kreitler et al., 2015).

| C A S E S TUDY: OPP ORTUNITIE S FOR CON S ERVATI ON OF CLIMATE REFUG IA AND C ARBON UNDER THE NORTHWE S T FORE S T PL AN
The US planning processes with the greatest potential for expanding protection across multiple jurisdictions typically have occurred when the desire to avoid imminent listing of a species as endangered or threatened results in strong political support for overcoming bureaucratic barriers to coordination. For example, the 2015 planning process for conservation of the greater sage grouse (Centrocercus urophasianus) resulted in new protections on 67,000 km 2 of habitat "strongholds" in 10 western states (Pidot, 2018).
The Northwest Forest Plan (NFP; Figure 7) provides another example which exhibits the two key requirements for coordinated multi-agency conservation planning (Spies et al., 2019). Strong political impetus was generated when a judicial decision upended the status quo of timber management on federal lands, finding that it failed to conserve the northern spotted owl (Strix occidentalis caurina), whose declining population trends were related to the loss of older coniferous forest habitat (Noon & Blakesley, 2006  Additional areas under conservation management may be necessary to achieve the NFP's conservation goals (Dunk et al., 2019), demonstrating that the 30% threshold is insufficient in some contexts ( Figure 1).
The NFP region exemplifies the three challenges to adaptation and mitigation planning described previously. Planners must reconcile a diversity of data on the location of climate refugia.
There are both commonalities and contrasts between adaptation priority areas as represented by refugia for rare and endemic old-forest-associated taxa , songbird and tree species refugia , potential topographic microrefugia  and climate connectivity areas . Starplots of adaptation and mitigation values for the NFP region ( Figure 7b)  The NFP region exhibits in microcosm the geographic contrast between adaptation and mitigation priority areas that is also evident at the continental extent. Priority areas for carbon protection are found predominantly in the northern coastal section of the NFP region, where carbon-releasing fire disturbance is infrequent (Buotte et al., 2020). In contrast, the southern portion of in the NFP region hold greater importance for species diversity and species refugia ( Figure S1) . The original 1994 Northwest Forest Plan addressed representation issues by distributing latesuccessional reserves across the region's physiographic provinces (outlined in Figure 7a). The NFP revision process would need to address whether adaptation and mitigation goals would be similar representative, or whether mitigation goals should primarily focus on wet coastal forests.

| CON CLUS I ON: TOWARD B E T TER INTEG R ATI ON OF CLIMATE ADAP TATI ON AND MITI G ATI ON G OAL S INTO PROTEC TED ARE A NE T WORK E XPAN S I ON
A necessary foundation for effective biodiversity-focused adaptation and nature-based mitigation policy is a clear national strategy that is linked to agency mandates and regional planning processes.
Such a strategy should address the issues identified above regarding the availability and appropriate use of spatial data, reconciling management for adaptation and mitigation, establishing and coordinating national and regional targets, and tracking progress toward target achievement.
Notwithstanding some encouraging progress to date sparked by Canada's Target 1 process, prioritizing new areas for protection faces numerous barriers in achieving biodiversity, climate adaptation, and mitigation goals. These range from data availability limitations to societal pressures, including economic opportunity costs of conservation. Achieving such outcomes will require that more explicit policy links be made between Canada's commitments under the CBD and the Paris Agreement, although some promising steps are being made with the recent issuance of a new climate plan that acknowledges the essential roles of nature and of Indigenous-led conservation (ECCC, 2020c). Funding decisions so far have been substantially opportunistic (i.e., based on submitted proposals), rather than through proactive and systematic prioritization of conservation targets underrepresented within the existing protected area network. Although potential projects are widely distributed across Canada, most newly protected land has to date been located in a few regions (ECCC, 2020b). A process is needed that provides transparent evidence-based evaluation of where protection of biodiversity, climate refugia, and carbon storehouses can be aligned.
Moreover, even if both quality and quantity area-based targets are met, biodiversity conservation will not be achieved without concerted attention to limiting the extent and intensity of development within unprotected areas (Ray et al., 2021). Efforts in the US as in Canada would benefit from development of a national biodiversity-focused adaptation and naturebased mitigation strategy which described a nested sequence of analyses that assess how regional actions can best contribute to achieving national targets and ensuring connectivity and complementarity among regions (Law et al., 2004). In this essay, we have provided examples of such complementary national (Figures 5 and 6) and regional-scale (Figures 4 and 7) assessments. This proposal aligns with other recent mitigation-focused proposals, such as for a National Strategic Carbon Reserve to prioritize retention of ecosystem carbon on federal lands, especially within moist forests in the US Pacific Northwest and Alaska (Dellasala et al., 2020).
Greater coordination of energy and natural resource production, mitigation, and climate adaptation policies in whole-of-government approaches, including re-direction of subsidies for fossil fuel and natural resource development, is also essential (Ray et al., 2021).
At a national level, such coordination could involve elimination of incentives for extractive energy production in climate refugia and carbon-rich ecosystems, especially on public lands. Measures such as a fossil fuel leasing moratorium on public lands would limit both downstream effects of fossil fuel use and direct emissions during the extraction process (Eilperin & Grandoni, 2020). Growth-inducing linear infrastructure, for example, pipelines and basin-opening roads, also merits scrutiny outside of project-specific environmental impact assessments (Johnson et al., 2020).
National biodiversity-focused adaptation and nature-based mitigation strategies should be integrated with other pathways for conservation management. In Canada, recovery planning mandated by the Species At Risk Act for widely distributed species of concern such as caribou (Rangifer tarandus) presents an opportunity for achieving co-benefits from protection of refugia and ecosystem carbon (Wells et al., 2020). In the US, regional species conservation processes, such as those described above involving the sage grouse and northern spotted owl, should be broadened to consider adaptation and mitigation goals, and tiered to goals presented within the national strategy.
The Fish and Wildlife Service (FWS), the US agency responsible for conservation of terrestrial endangered and threatened species, not only manages land directly, but also consults with other land management agencies when the latter's actions may affect species of concern (Jeffers, 2008). Such "Section 7" consultations provide a potential pathway to advance coordinated protection of climate refugia across multiple jurisdictions. Designation of "critical habitat" for threatened and endangered species by FWS provides another pathway for refugia protection. However, recent policy changes (FR 85 82376) and case law (Weyerhaeuser Co. v. U.S. Fish and Wildlife Service, 17-71 (Nov. 27, 2018)), which prohibit FWS from designating critical habitat in an area where the species of concern is not currently present, are problematic in that they may prevent protection of areas which are newly suitable for a species due to shifting climates.
Private and non-federal public lands harbor a significant proportion of North America's biodiversity (Rosa & Malcom, 2020) and climate refugia (Figure 6b). Federal governments can also use funding to incentivize participation by non-federal landholders. Support of Indigenous-led protected area proposals by Canada's Target 1 Challenge Fund provides an example of the potential impact of such programs (ECCC, 2020b). In the US, protection of refugia on nonfederal lands can be achieved via the Land and Water Conservation Fund, which provides matching funds to state and local authorities, and the Conservation Reserve Program (CRP), which pays private landowners to manage land for conservation (Szentandrasi et al., 1995). US counties with the highest proportion of CRP lands or conservation easements show high values for conserving climate connectivity and above-ground carbon (Figure 6b). US states such as California which have endorsed area-based protection targets and mitigation goals could also benefit from developing a coordinated adaptation and mitigation strategy that builds on existing State Wildlife Action Plans.
In 2009, the US Department of Interior established a network of 22 regional Landscape Conservation Cooperatives (LCC) in order to coordinate efforts by federal agencies, states, tribes, and nongovernmental organizations to address climate adaptation and other broad-scale conservation issues (Jacobson & Robertson, 2012).
However, LCC funding was terminated in 2019 by the subsequent federal administration, and although Canadian government staff and other experts participated in transboundary LLCs, funding from Canada was absent. Federal science agencies can play a key role in supporting biodiversity-focused adaptation and nature-based mitigation efforts by reviving and strengthening the network of Landscape Conservation Cooperatives and other "communities of practice" (Wenger, 1999) that bring together researchers and practitioners (e.g., staff from governmental agencies, First Nations, and NGOs). Our

ACK N OWLED G EM ENTS
The manuscript benefited from comments by L. Bell, H. Cooke, M.
Reid, and W. Vanasselt. The Wilburforce Foundation provided support for C.C. as part of the AdaptWest Climate Adaptation Planning Project (https://adapt west.datab asin.org).

CO N FLI C T S O F I NTE R E S T
The authors declare no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
CC conceived the study; CC and JR wrote the paper.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data presented in this study are openly available at adaptwest. databasin.org.