Introduction
Climate change is creating a warmer, more acidic, and oxygen-limited ocean, trends that are expected to continue at least until 2050, even under the most ambitious emission reduction scenarios (
IPCC 2021). These climatic changes affect marine species through many pathways, including physiological processes (
Pörtner et al. 2008;
Pörtner and Peck 2010) and the spatial/temporal dynamics of ecosystems that they inhabit (
Molinos et al. 2016;
Scheffers et al. 2016;
Henson et al. 2017). Such impacts alter the abundance and distribution of species, ecosystems, fisheries, and livelihoods of those who depend on them (
Lotze et al. 2019;
Boyce et al. 2020;
Bryndum-Buchholz et al. 2022). Emission mitigation is the most straightforward path to reducing climate impacts (
Bates et al. 2019;
Fawzy et al. 2020), but marine protected areas (MPAs) are a vital adaptation approach that can support climate resilience (e.g.,
Roberts et al. 2017;
Jacquemont et al. 2022). In Canada, MPAs are a legislative management tool established under the
Oceans Act, used by the government to protect ocean spaces for long-term biodiversity conservation through targeted restrictions on activities based on risk to the conservation objectives associated with a given location (Oceans Act S.C. 1996 c. 31. 2018). Within it, an MPA is defined as “
a clearly defined geographical space,
recognized,
dedicated and managed,
through legal or other effective means,
to achieve the long-term conservation of nature with associated ecosystem services and cultural values” (
DFO 2010). In addition to MPAs, Canada’s marine conservation strategy includes other regulatory tools, including “other effective conservation measures” (OECMs), which represent spatially-defined areas that provide long-term, measurable biodiversity conservation benefits, including National Parks and National Marine Conservation Areas (NMCAs; administered by Parks Canada), Migratory Bird Sanctuaries and National Wildlife Areas (NMWA; administered by Environment and Climate Change Canada), and Marine Refuges (administered by Fisheries and Oceans Canada). Although not necessarily or formally designated for biodiversity protection, OECMs represent measures that confer biodiversity conservation benefits and contribute to Canada’s international conservation commitments (
Bryndum-Buchholz et al. 2022). For example, “Marine Refuges” are areas regulated under Canada’s Fisheries Act, primarily for fisheries management purposes. However, these areas also provide biodiversity conservation benefits by limiting activities that could otherwise negatively impact marine ecosystems. The Marine Protection Standard (
DFO 2023) provides a standardized framework for permissible activities within federal MPAs, restricting activities such as oil and gas development, offshore mining, and bottom trawling, aligning with IUCN standards (
IUCN-WPCA 2018). While the protection criteria for OECMs share similarities with those of MPAs, they utilize a “case-by-case” method to evaluate the influence of various economic activities on their conservation objectives. For Marine Refuges, this process is applied when establishing regulations for activities not addressed by the Fisheries Act that conflict with biodiversity conservation objectives. Hereafter, we refer to MPAs and OECMs collectively as “MPAs”, due to their joint contributions to Canada’s conservation targets (e.g., “30 by 30”).
The ability to adapt to changing conditions becomes particularly important for static conservation measures in a dynamically changing climate. However, marine conservation plans rarely consider climate change and often lack tangible design and management actions to mitigate its impacts (
Wilson et al. 2020;
O'Regan et al. 2021;
Bryndum-Buchholz et al. 2022;
Bryce and Hunter 2024). For instance,
O'Regan et al. (2021) reported that climate change was considered in only 26% of reviewed Canadian MPA management plans compared to 57% in the United States. While some management plans for Canadian MPAs have reviewed potential climate change-related indicators that could be monitored (e.g.,
Faille et al. 2019), most are site-focussed, abstracted from the broader network where climate change impacts may be most apparent. Thus far, Canada’s growing marine conservation network (CMCN) has no explicit objective for addressing climate change and/or supporting climate resilience (
Bryndum-Buchholz et al. 2022). However, in 2021, the Fisheries and Oceans Canada mandate outlined directives for modernizing the Oceans Act. This included considering climate change impacts on marine ecosystems and species in regional ocean management, ensuring measurable indicators and objectives, and expanding climate vulnerability work to inform conservation planning and management (
Prime Minister of Canada 2021).
Incorporating climate change vulnerability into marine conservation decision-making aligns with scientific recommendations for protections distributed across a spectrum of anticipated changes or “climate risks” (e.g.,
Foden et al. 2019;
Wilson et al. 2020). By incorporating and equitably representing anticipated levels of change into conservation measures for a given planning region, a network of conservation areas will be better equipped to support biodiversity over the long-term. Areas of rapid change can be hotspots for emergent new ecosystems and/or may represent future areas of importance for species of conservation concern, biodiversity, or productivity if protected from other stressors (
Cheung et al. 2009;
Rilov et al. 2020;
Queirós et al. 2021). Alternatively, placing protections in locations where change is slow can provide refugia for species with lower or narrower temperature tolerances and safeguard areas where species turnover may be commensurately lower (
Queirós et al. 2016;
Rilov et al. 2020;
Arafeh-Dalmau et al. 2021). As
Wilson et al. (2020) summarized, it is widely suggested that protecting the full range of climate impacts can support connectivity and/or seeding of shifting populations, ecosystem functions, representative habitat types, and adaptivity of species to environmental change. In this respect, resilience is not limited to increasing the ability of protected ecosystems to resist change (e.g.,
Jacquemont et al. 2022) but also safeguarding biodiversity and ecosystem functions as they evolve.
Canada has made substantial progress towards its 2030 marine conservation targets, with just under 15% of its exclusive economic zone (EEZ) protected as of 2023. In the next six years, Canada has committed to establishing MPAs in another 15% of its EEZ, comprising ∼862 000 km
2, equivalent to 20% of its land area. The conservation objectives of the CMCN prioritize long-term protection of marine biodiversity, ecosystem function, and special natural features (
Government of Canada 2011). While regulatory measures like MPAs and OECMs offer enduring protection, the planning of conservation areas at the bioregional and national levels has, for the most part, not directly considered how conservation effectiveness might evolve over time (e.g.,
DFO 2014,
2018). While some networks have begun to assess the impact of climate change on their conservation strategies (
Lewis et al. 2023;
Rubidge et al. 2024), overall, there remains uncertainty in this area. Tools such as climate change vulnerability assessments are available to help bridge this gap, providing the information needed to design representative, climate-resilient conservation networks in Canada. Vulnerability assessments provide information on how species and/or ecosystems are anticipated to respond to climate change and have been recommended to help conservation planners create more representative, climate-resilient networks, which can support a multitude of future conditions (
Wilson et al. 2020;
Bryndum-Buchholz et al. 2022).
In this study, we evaluated the representation of climate change vulnerability within the CMCN to identify opportunities for supporting its climate resilience. Our analysis covered national, ocean basin, and bioregional scales, and focused in on a draft conservation area network in Eastern Canada (
Fig. 1). We employed the spatially explicit Climate Risk Index for Biodiversity (CRIB) developed by
Boyce et al. (2022), which calculates climate change vulnerability scores for marine species based on how their traits interact with historical and projected ocean warming, and other factors such as their conservation status and exposure to other stressors (
Fig. 2). Averaging the individual species vulnerability scores at each location yields ecosystem-level vulnerabilities that characterize the emergent likelihood that biodiversity (e.g., the species assemblage at a given location) will be adversely impacted, locally extirpated, or potentially extinct due to climate change. A higher vulnerability score denotes an ecosystem containing many species expected to experience climate-driven population decline, geographic distribution shifts, thermal stress, and drastic community restructuring, with greater magnitude and rapid onset impacts. In short, we expect such ecosystems to be rapidly rewired under climate change within the given projection timeframe. By comparing CRIB-derived vulnerabilities within MPAs to surrounding areas, we identified representativity gaps and potential issues associated with a climate change-agnostic network design approach. Our findings highlight the importance of incorporating climate resilience into the CMCN as Canada works towards its 30% marine conservation target.
Discussion
Our results indicated that the Canadian EEZ, throughout the Pacific, Arctic, and Atlantic Oceans, exhibited generally low climate change vulnerability (<0.5) compared to the global picture, where it ranges from 0 to 1 (
Boyce et al. 2022). A notable exception to this was the Strait of Georgia, where vulnerability in some areas approached 0.64 under the high emissions scenario (
Fig. 4). However, in all cases, average climate change vulnerability inside and outside of MPAs increased under the high emissions scenario, suggesting that the magnitude of climate change vulnerability in Canadian marine ecosystems is tied to the extent of emissions mitigations implemented globally.
One strategy to support ecosystem resilience in a regional planning area is to ensure representative protections across a broad spectrum of habitat types, oceanographic features, and species assemblages. Representativity is a design feature advised by the Convention on Biological Diversity (
CBD 2010) and adopted by nations worldwide (
Balbar and Metaxas 2019), including Canada (
Government of Canada 2011). With a changing climate, representativity takes another dimension, where biodiversity protection achieved using spatial conservation measures becomes a moving target. When appropriately distributed, scaled, and protected, conservation networks have the potential to support transient ecological states as conditions continue to change in a fashion akin to terrestrial protected area “corridors” for shifting distributions and species turnover (
Rice and Houston 2011;
Fredston-Hermann et al. 2018). In this sense, ensuring the resilience of spatial conservation measures involves representing a broad spectrum of climate vulnerabilities, safeguarding not only slow-changing areas (or “climate refugia”) but also those undergoing more rapid changes. This strategy “hedges” against uncertainty in ecosystems' responses to climate change, increasing the likelihood that spatial conservation measures aimed at maintaining biodiversity will remain effective even as conditions change (
Queirós et al. 2016;
Fredston-Hermann et al. 2018). Additionally, it protects emerging ecosystems, some of which may become highly productive and diverse (
Queirós et al. 2016;
Queirós et al. 2021). As Canada advances its Marine Conservation Targets and fulfills the objective of designing ecologically comprehensive, resilient, and representative conservation networks, it is crucial to consider how ecosystems are likely to change to ensure that spatial conservation measures are robust to anticipated change (
Bryndum-Buchholz et al. 2022).
High climate representativity (>75%) in all but one (Arctic Archipelago) of the 12 bioregions indicates that the distribution of MPAs in the CMCN proportionally represents most of the modelled climate change vulnerability, even when emissions scenarios are uncertain. However, there was some variation among bioregions and between emissions scenarios. Under the high emission scenario, climate change vulnerability was higher (
Fig. 4) and representativity of bioregional climate change vulnerability inside of MPAs was improved relative to the low emissions scenario throughout the Western Arctic, Hudson Bay Complex, Southern Shelf, Arctic Basin, Strait of Georgia, and Newfoundland-Labrador Shelves bioregions, (
Fig. 5A). Given the correspondingly high vulnerability scores of these bioregions relative to others in the EEZ, this improved representativity under increased warming could be indicative of broad homogenization of habitats (e.g., a transition to ice-free habitats). This could foreshadow the immigration of new assemblages seeking highly altered, warmer areas (e.g.,
Cheung et al. 2011;
Alabia et al. 2020). However, with representativity levels still below 80%, there remains potential to improve the representation of higher vulnerability areas.
Conversely, climate representativity of bioregional climate change vulnerabilities inside MPAs was substantially reduced under the higher emissions scenario in the Arctic Archipelago, Eastern Arctic, and Offshore Pacific bioregions (
Fig. 5A). In these bioregions, rates of environmental change have been shown to vary between different habitat types. For instance, in the Eastern Arctic, sea ice has been melting less rapidly than in other areas in the Arctic Ocean (
Post et al. 2013). In the Offshore Pacific, complex oceanography and proximity to an ocean warming hotspot have led to predictions of rapid and spatially variable ecological change (e.g.,
Okey et al. 2014). In the Arctic Archipelago, the complex geomorphology of the bioregion's constituent oceanic islands may contribute to variable responses to climate change. This variability could account for the observed trend of decreasing representativity under the high emissions scenario, whereby changes are more pronounced and heterogeneous than other bioregions. Bioregional differences in ecosystem response to warming highlight the complexity of conservation planning in a changing climate. Nonetheless, by understanding these differences, conservation planning exercises can help make informed decisions and identify areas that may require adaptive management strategies.
Where gaps existed in climate representation, it was often because the placement of MPAs was skewed towards areas with lower vulnerability relative to the rest of the bioregion in question (
Figs. 3 and
4). This is a common theme in marine conservation, where conservation planners prioritize both biological and human-use objectives, thus avoiding areas of high human usage to limit socio-economic costs/trade-offs (e.g.,
DFO 2018). These high-human-use areas, such as fishing grounds, are typically closer to shore (
Halpern et al. 2008), where waters are shallower and often more influenced by atmospheric warming (
IPCC 2018;
Albouy et al. 2020). Gaps in coverage of high vulnerability areas suggest that the representativity of regional biodiversity could decrease over time, particularly in nearshore environments, as novel assemblages and ecosystems emerge with a changing climate. Increasing the designation of MPAs in higher-vulnerability areas could address this representativity gap, helping to ensure that the conservation of biodiversity in networks is more resilient to change (
Cheung et al. 2009;
Bryndum-Buchholz et al. 2019;
Descamps and Strom 2021). This approach could also contribute to distributing protection from other stressors, such as fishing and fossil fuel drilling, and support areas poised to become more productive. Indeed, it has been predicted that bioregions in the Arctic Ocean, some of which had lower levels of climate representativity (e.g., Arctic Basin and the Arctic Archipelago,
Fig. 5A), will experience the most significant transformation of species assemblages of the entire global ocean, with a large magnitude of new species moving into this area, high turnover of ecosystem structure (
Alabia et al. 2020;
Florko et al. 2021;
IPCC 2023), and increasing primary productivity (e.g.,
Wassmann 2011), potentially leading to an increase in catch potential of many existing and incoming commercial species (
IPCC 2018;
Tia et al. 2019).
Reflecting the widely observed tendency toward the biased designation of MPAs in low-vulnerability areas and the low representativity of conservation areas in traditionally high-use coastal areas, MPAs were disproportionately designated in deeper waters across all three ocean basins, particularly in the Pacific (
Fig. 7). While this has led to climate representativity gaps throughout the EEZ, it also highlights a representativity gap for shallow-water coastal and continental shelf habitats, and their associated species assemblages and ecosystem functions.
Coastal shallow-water habitats, such as eelgrass (
Zostera marina) and kelp (
Laminarian spp.) beds, play crucial ecological and economic roles by supporting keystone species and providing essential ecosystem services like carbon sequestration and coastal protection (
Lilley and Unsworth 2014;
Röhr et al. 2018;
Infantes et al. 2022). These areas are also hotspots for commercial fishing and thus face cumulative stressors that reduce their resilience to environmental changes (
Halpern et al. 2008;
Sharples et al. 2013;
Beauchesne et al. 2020). In contrast, deep-sea habitats host slow-growing, late-maturing species like octocorals and stony corals, which are highly sensitive to disturbance from fishing activities and emerging threats such as deep-sea mining (
Gollner et al. 2017;
Beazley et al. 2021;
Paulus 2021). As fishing efforts expand into deeper waters, concerns mount over increased impacts on these ecosystems, underscoring the need for comprehensive protections to sustain their fragile biodiversity and ecosystem functions.
Given the varying ecological functions and vulnerabilities of shallow and deep-water ecosystems to climate change, it is essential to prioritize balanced protection across both. This study underscores the need to increase protections, particularly in shallow-water coastal areas. Prioritizing representativity across different depths, habitats, and climate change vulnerabilities within the CMCN can enhance biodiversity protections and promote ecological stability amid environmental changes (
McLeod et al. 2009;
Pettersen et al. 2022).
Our analysis of a draft conservation network proposed for the Scotian Shelf-Bay of Fundy Bioregion supports this approach. Prioritizing representativity of habitats (including depths and sediments), oceanographic features (e.g., temperature regimes), and geography (i.e., targets stratified for the Eastern and Western Scotian Shelf) in the conservation planning process (
DFO 2018) resulted in balanced protection of regional diversity and high climate representativity in the bioregion (
Fig. 6). This study strongly advocates for maintaining representativity as a cornerstone in conservation planning efforts as Canada progresses towards its conservation targets, ensuring the long-term resilience of biodiversity in marine ecosystems, whether embodied by present day or future/emerging ecological communities (
Queirós et al. 2016,
2021).
Caveats
While integrating climate change considerations into conservation planning is crucial (
Tittensor et al. 2019;
Bryndum-Buchholz et al. 2022), it is essential to acknowledge the inherent uncertainty associated with available tools and data products. Predictions from global ocean climate models may introduce uncertainty at the local scale, potentially leading to over- or under-representation of species-specific vulnerability scores in certain areas. However, general ocean-scale patterns, such as the poleward movement of isotherms, and ecosystem-level vulnerabilities can provide insight into broader trends expected in a warming ocean. Moreover, considering the median conservation area size of ∼100 km
2 in the CMCN, the resolution of the CMIP oceanographic models used in this study and by
Boyce et al. (2022) are likely sufficient to resolve expected gradients of change within planning regions and MPAs in Canada.
Further, while broad-scale physical ocean models predicting future thermal conditions are more readily available (e.g., the CMIP model family), they may not adequately capture other environmental conditions such as oxygen concentration and pH level. Consequently, climate vulnerabilities may be underestimated, particularly those related to oxygen reduction in deep areas (e.g.,
Beauchesne et al. 2020;
Ross et al. 2020). This can lead to further habitat losses and prevent species from migrating deeper to avoid warming surface waters (
Stortini et al. 2017;
Thompson et al. 2023). Additionally, the CRIB does not account for potential shifts in interspecific interactions (e.g., predation shifts
Zabihi-Seissan et al. 2024) that could result in higher ecosystem-level vulnerabilities.
Recognizing the potential underestimation of climate vulnerabilities, taking a precautionary approach is crucial, ensuring that networks offer adequate protection and are broadly representative to support long-term resilience. While a representative approach to conservation planning is recommended based on our results, this is only one dimension of effective, climate-resilient network design. Conservation planning and design principles must also consider spacing between MPAs, population connectivity (
Beger et al. 2022), and fisheries management objectives (
Fovargue et al. 2018).
Notwithstanding these limitations, this study used models, projections, and data to assess Canada’s MCNs based on the best available decision-support tools. Importantly, inferences drawn from this work provide a needed initial foundation on which to start the conversation regarding building climate resilience into Canada’s national marine conservation network.