Social–ecological connections across land, water, and sea demand a reprioritization of environmental management

Increasing rates of environmental change, the crisis of biodiversity, the loss of ecosystem services (ES), and the risk of surprises from tipping points (nonlinear ecological or social transformations) highlight the need to find different approaches to navigate society toward ecological sustainability (Vitousek, 1997; Scheffer et al., 2001; Rockstrom et al., 2009; Lindenmayer et al., 2010; Carpenter et al., 2015; Organisation for Economic Co-operation and Development, 2017; Filbee-Dexter et al., 2018). Despite many sectors of society striving for sustainability and balanced interactions with the environment, the current focus and effort is insufficient to generate sustainable solutions because it does not account for the whole problem. The issues that humanity collectively needs to address are relevant both to interventions that react to problems and those that seek to prevent them. Here, we explore how a focus on the social and ecological connections and feedbacks (both positive and negative) across ecosystem domains (land, freshwater, and marine) can promote resilience to multiple future threats (Biggs et al., 2011; Selkoe et al., 2017; Lenton, 2020).

Ecological and social knowledge accumulated over decades has highlighted the role of biophysical subsidies and connectivity across a “hilltops to ocean” continuum (Polis and Hurd, 1996; Ramesh et al., 2015; Gounand et al., 2018). Despite this, management practice still tends to be isolated by ecosystem domain (Singh et al., 2021; Threlfall et al., 2021), often using different approaches and frameworks, with inequities in data and knowledge (Figure 1). Land can be privately owned, but the ocean is usually in the public domain, and this discrepancy leads to different aspirations and targets (Figure 1B). Environmental management on land is targeted at species (animal or plant) or habitats and focused on the ES that biodiversity provides, while in the ocean, resource extraction is prioritized (Figure 1C and D). The relative visibility of changes and ease of access to the 3 ecosystem domains also drives differences in social awareness and how we allocate science and management resources. Consequently, environmental data quality and quantity tend to be greater in volume, velocity, and variety for land than in freshwater, with both well ahead of marine (Figure 1A).

Systems thinking and more holistic management approaches such as ecosystem-based management are much discussed concepts but have proven difficult to implement (Christensen et al., 1996; Ruckelshaus et al., 2008; Granek et al., 2010; Thrush and Dayton, 2010). Because environmental issues are often dealt with in isolation, governments, businesses, and individuals can fail to first identify and then act on the connections that lead to and prevent abrupt unexpected state changes or tipping points. Tipping points occur when a system’s environmental and social stressors (or phenomena) have intensified to the point where a system shifts to a different state (often for the worse from a human perspective). Further sudden (nonlinear) changes from that point are also possible (i.e., multiple tipping points can exist), and state shifts can trigger other state shifts in distant locations, for instance, via nutrient and water flow (Rocha et al., 2018). There is a growing realization that interactions among social and ecological components of ecosystems are crucial for fostering positive environmental outcomes where tipping points can be foreseen and prevented. For example, an extensive literature and multiple cultures highlight a deep interconnection between social and ecological systems that coevolve across space and time, allowing adaptation in times of change (Folke, 2006; Carpenter et al., 2015; Osterblom et al., 2017; Filbee-Dexter et al., 2018; Nystrom et al., 2019; Yletyinen et al., 2019).

In a highly connected world with the certainty of climate change and further anthropogenic exploitation (Nystrom et al., 2019), it is imperative to explore and implement robust management regimes. To be robust to a range of plausible futures, management should traverse land, freshwater, and marine ecosystem domains and consider their connectivity in important functions, flows, feedbacks, and impacts (Schiel and Howard-Williams, 2016). We refer to these critical connections as “cross-domain connections,” and “connectivity” as the movement of materials, energy, ideas, and the expansion of social structures and practices across land, freshwater, and sea. We focus on these cross-domain connections as these represent major opportunities to improve environmental management and governance. While the level of integration of hierarchies in governance may vary both within and across domains (Singh et al., 2021), our focus is on the ecological and social linkages. This approach and framing are consistent with many Indigenous Peoples' worldview, knowledge, and practice (e.g., McGregor, 2004; Clapcott [Ngāti Porou] et al., 2018). Here, we lay out the latent opportunities for management, which currently does not adequately acknowledge these cross-ecosystem domain connections, and therefore has limited potential to enhance sustainability and the resilience needed to prevent undesirable tipping points.

The ideas presented in this perspective piece evolved from a workshop designed to address the siloed nature of the New Zealand National Science Challenges, which arguably reflect broader management and governance structures in New Zealand and globally. These National Science Challenges were deliberately focused on separated environmental domains (e.g., marine, freshwater, and land). Our collective concern was that the connections between ecosystems were being ignored, or at least not prioritized, thus limiting research and solution sets (e.g., management interventions). The workshop involved an internationally diverse set of 17 researchers whose research spans the 3 ecosystem domains and multiple disciplines including ecology, social sciences, economics, Māori and indigenous studies, natural resource management, and systems modelling. Based on our collective experiences and discussions during the workshop, we (1) present our perspectives on the social–ecological properties of cross-domain connections that emphasize the need for management reprioritization, (2) demonstrate that the strength and speed of cross-ecosystem domain feedbacks in the social (management) versus ecological components often differ (illustrated with a simple dynamic model to demonstrate possible environmental outcomes associated with delayed management responses [relative to the generation of ecological impacts]), and (3) offer guidance for cross-domain environmental management and research priorities that aim to better prepare for and mitigate tipping points. Throughout our narrative, we connect broader concepts to 3 case study examples (Boxes 1–3). Importantly, we highlight the benefits of investing effort and focus on cross-domain and cross-scale (space and time) connections to identify the common threads that support opportunities for change.

Humanity has known of cross-domain connections for centuries but appears to be continually surprised by their consequences. For example, damming rivers in southern California reduced sediment transport to the coast, enhancing beach erosion (Sherman et al., 2002). Extraction of water for land irrigation in Australian rivers altered estuarine and lagoon circulation regimes, creating complex geochemical and physical feedbacks resulting in hyper-eutrophic states (Laurance et al., 2011). Upstream damming, sand mining, and enhanced erosion of riverbeds have led to saltwater intrusion of productive lands in the Mekong River Delta, with massive economic, social, and ecological consequences (Eslami et al., 2019). However, these unintended effects are not always deleterious for ecological sustainability. For example, the management of invasive predators on islands created unintended positive feedbacks, where removal of predation pressure on seabird populations (and their subsequent recovery) resulted in increased inputs of marine sourced nutrients, stimulating forest growth (Fukami et al., 2006).

While our current environmental management practices have had some successes (e.g., reduction of DDT pesticide in the Baltic Sea and air pollutants and acid rain in Europe and North America; Helsinki Commission [HELCOM], 2010; Grennfelt et al., 2020), global trends show that this is not sufficient to prevent tipping points. Rather than being trapped in the narrative that science needs to better inform management, we expand aspirations and view the issues through a cross-domain lens to suggest a reprioritization of management efforts. Considering even simple social–ecological properties of ecosystems through this lens can lead to improved environmental management, which may help to break the cycle of sectarian governance that currently facilitates blame and inaction (Howlett, 2014). These simple but fundamental social–ecological properties (SE-P) are as follows:

Generally, water flows from land to sea, meaning that the aquatic environment is the net recipient of change on land, and ultimately, the coastal marine environment is the net recipient of many changes in both land and freshwater ecosystems (Figure 2). This downstream connectivity has resulted in degradation of aquatic environments globally, which is perpetuated by disparities in ownership and accountability, governance, environmental targets, perspectives, data, management, and understanding of ecological processes between domains (e.g., see Boxes 1–3 for real-world examples of these dynamics playing out across domains; and Figure 1). “Water flows downhill” is a simple axiom, and yet current environmental management of terrestrial and freshwater ecosystem domains tend to devalue the considerable distal effects of land-based stressors in receiving aquatic environments (exemplified in Box 3’s critique of stressor limit setting; Figure 1F). There is a critical need to strengthen the social feedback that links downstream issues to upstream activities to sustainably manage both beneficial (e.g., subsidies) and detrimental (e.g., pollution) flows that are critical to biodiversity, and ecosystem functions and services in all 3 ecological domains (e.g., Boxes 1 and 2).

People can be separated from the consequences of their actions by the physical segregation of ecosystem domains and governance structures (Figure 2). This separation becomes a problem when the benefits that an individual accrues from an ecosystem are diminished by activities in another domain, where the people creating the impacts are separated from both those managing and those affected by impacts (depicted in Figure 2). A decoupling of decisions from the impact location reduces the feedbacks that would change management practices (e.g., political pressure to stop a particular activity; DeFries and Nagendra, 2017). Part of this issue stems from socially constructed boundaries around physical areas and jurisdictions (e.g., privately vs. publicly owned land; regionally vs. nationally managed areas; Brunson, 1998; see Box 2 for a discussion of values, ownership, and management scales with respect to wetland ecosystems). The perception of boundaries is not the same in the 3 ecosystem domains, leading to disparities in how people defend and protect physical areas (e.g., land, and sometimes wetlands and streams, are often in private ownership, and the ocean, and the resources it provides and sustains, are publicly owned; Figure 1B). People are more likely to adjust their actions to protect a forest or stream on their own property, or in a public space that is near to where they live, than a river or estuary that is some distance away (i.e., the psychological distance effect; Perry et al., 2021). Ecological processes transcend boundaries constructed by humans (DeFries and Nagendra, 2017), and this fractured social–ecological dynamic creates barriers to identifying and acting on the drivers of change and solutions.

The fracturing of ecosystem management across domains creates a situation where single environmental drivers (often within one ecosystem domain) are the focus of environmental impact mitigation (see Box 3 on stressor limit setting). Policy that targets specific ecological domains ignores the abundant literature demonstrating the flow of species, resources, and environmental effects between habitats within a domain (Frost et al., 2016), and between land and freshwater (Polis and Strong, 1996; Knight et al., 2005; Bartels et al., 2012), land and marine (Polis and Hurd, 1996; Sanchez-Pinero and Polis, 2000), and freshwater and marine domains (Palumbi, 2003; Wipfli et al., 2003; Gounand et al., 2018). A siloed domain focus can make it hard to identify drivers of change (see, e.g., Boxes 1–3), and this can leave humanity unprepared for future changes that arise from multiple (often nonlinearly interacting) stressors that originate in connected ecosystem domains (Sala et al., 2000; Crain et al., 2008; Darling and Côté, 2008). For example, land management for conservation remains largely focused on focal species (typically vertebrates) or on parcels of land demarcated by the dominant ecosystem type (e.g., a forest) rather than ecosystem processes and the potential for different habitats to be connected to other ecosystems (Figure 1C, E, and F; Kortetmäki et al., 2021). Freshwater management is typically conducted at a catchment scale and explicitly considers the linkages between land and freshwater domains (e.g., nutrient and sediment flows) with a focus on processes (e.g., catchments and upstream effects; Rouse and Norton, 2016). Land and freshwater management drive change in wetlands (Box 2), but it does not account for how wetlands can buffer change in estuaries or be impacted by salination associated with sea-level rise. Further, since freshwater quality and quantity are the main priorities, there is little consideration of any downstream effects in the marine domain. Management of the coastal marine environment usually recognizes the connectivity and impacts of decisions on land, but there are significant time lags and barriers associated with mitigating these distal effects (Figure 1F; Box 1; Schiel and Howard-Williams, 2016; Osterblom et al., 2017).

Ecological thresholds in one ecosystem domain often do not apply to other domains, and hence, a focus on stressor limit setting is unable to prevent threshold responses and tipping points (see Boxes 2 and 3). For example, the ecological limits for land-based sediment and nutrients into freshwater ecosystems are irrelevant for assessing the coastal marine environment’s response to these stressors. Being blind to the importance of understanding ecosystem responses and processes that drive differences in responses between domains has resulted in ecosystems passing tipping points (Rocha et al., 2015; Hicks et al., 2016; e.g., Boxes 1–3). Once ecosystems pass these tipping points, complex feedbacks and interactions often lock them in an undesirable state for indefinite periods and recovery can be impossible or extremely slow. For example, the slow accumulation of land-based nutrients and the severe eutrophication in the coastal waters of the Baltic Sea have occurred over centuries, but the benefits of management actions to curb the input of land-based nutrients will not be realized for decades due to the legacy effects in the system that slow recovery (Box 1). Often critical to these legacy effects are the slow growing ecologically important (e.g., habitat forming) species and associated biodiversity and ecological processes (Biggs et al., 2012; Andersen et al., 2017). Part of the problem with focusing on the stressor instead of the ecological processes (such as resilience and recovery dynamics) is that it eliminates the ability to generalize ecological responses spatially and temporally, or to identify interactive drivers of change and the legacy effects that result in recovery lags (Lindenmayer et al., 2010; Biggs et al., 2012).

There is no universal “right” scale for management, but identifying scale disparities between temporal and spatial, and social and ecological scales highlights a need for a different approach to environmental management. Scale in ecology has both spatial and temporal dimensions and reflects different levels of biological organization—individual organisms, populations, communities, and ecosystems (Levin, 1992). Scale in society and social science also has spatial and temporal dimensions in, for example, human relations, actions, governance, ownership, and politics (Clark, 1985; Gunderson and Holling, 2002; Cash et al., 2006; Cumming et al., 2006; Pyyhtinen, 2017). These structural scales of environmental management have political and social consequences that affect environmental processes and management. For example, management of environmental issues at a global (e.g., climate change) or a national scale (e.g., fisheries or pollution) can remove responsibility or power to act at a local scale even if the impacts are felt locally (Brashares et al., 2014; Haarstad, 2014). Further, these scale mismatches can create situations where actions to mitigate one environmental problem (e.g., demand for bioenergy for climate change mitigation) can yield negative consequences for other important aspects of the environment (e.g., land uses and biodiversity; Pörtner et al., 2021). Similarly, broad-scale acquisition of data to inform management can prevent adaptive responses to ecological processes that occur at local scales. The negative consequences of scale mismatches (Cumming et al., 2006) could be better addressed through local, place-based management that considers connections and linkages (Herse et al., 2020; Pörtner et al., 2021).

Ecosystems on land are often managed at local/regional scales, but marine ecosystems are managed at national to international scales. National to international scale governance disempowers local actors from effecting change or driving adaptation (Pisor et al., 2022) and instead places responsibility on governments or intergovernmental agencies, leading to, for example, differences in the way people extract resources from land versus sea (Singh et al., 2021). The issue is not the specific scale of management but rather in the scale mismatches between management and ecological and social processes (discussed in Boxes 2 and 3). However, there are also potential negative consequences of increasing connectivity in the social structures of governance. For example, some ecosystems maintain their healthy state through the application of rules and knowledge of local people, whose sustainable environmental governance systems may then be challenged or eroded by new governance approaches (Young et al., 2006; Longo, 2012), which tend to be at larger scales where people are separated from the consequences of their actions or management decisions (see SE-P2 above).

Recognizing connections across domains and scales offers opportunities to target social–ecological connections that could change the practices that currently obstruct aspirations of ecological sustainability and associated management actions. However, these opportunities only represent possibilities because the social–ecological context is complicated by the social constraints and path dependencies arising from existing policy and institutional frameworks (e.g., Box 1—Baltic Sea; Blenckner et al., 2015). Such critical impediments often sit at the interface between science, governance, and society (Thrush et al., 2016; Stenseth et al., 2020). The exceptions are responses to visible and immediate impacts such as whale stranding, oil spills, or wildfires. Oceans are possibly at greatest risk from scale mismatches due to 2 factors; first, the fallacy that oceans are too big to fail, having infinite capacity for recovery, and the ability to dilute and disperse contaminants, and second, that many of the effects on marine ecosystems are not in the public consciousness (due to a lack of visibility), and often arise from multiple stressors (Thrush et al., 2016; Selkoe et al., 2017). Common to all domains are impacts that generate immediate economic consequences (e.g., invasive species affecting industry), are highly visible (e.g., oil spills, severe eutrophication, land erosion, desertification, and wildfires), and elicit an emotional response (e.g., whale stranding), which show faster social feedbacks than slow insidious impacts exemplified by climate change and biodiversity loss. The slow and insidious impacts are often discounted as problems of the future but can lead to intergenerational injustice (Treves et al., 2018). This reactive and near-sighted management prioritization generates a focus on the short-term immediate impacts (e.g., the oil spills) rather than the chronic subtle cumulative effects on ecosystem components that have long-term effects because they affect slow to recover processes (e.g., over-fishing and terrestrial run-off into coastal waters remove key species and reduce regional biodiversity, which are generally very slow to recover). These slow changes also alter our perceptions of what ecological recovery looks like and the targets we set (i.e., shifting baseline syndrome; Soga and Gaston, 2018).

While the principles above may seem complicated and numerous from a management perspective, important insight can be gained by focusing on 2 critical elements that run through the principles and examples (Boxes 1–3): (1) the recognition of cross-domain connections as drivers of change and (2) the importance of protecting the slow-to-recover ecosystem elements (species and processes) that are often eroded by chronic and cumulative stressors (Heinze et al., 2021). There are examples globally of environmental management incorporating cross-domain connections (reviewed in Threlfall et al., 2021), and these examples provide insight. For example, Box 1 highlights some lessons in managing eutrophication in the Baltic Sea where collaborations across countries and agencies have begun to address the downstream effects of agriculture and urbanization on the Baltic Sea eutrophication status.

To illustrate the dynamic implications of slow management responses to ecological degradation exemplified by those outlined in Box 1, we present a simple control-theory model (Figure 3) to demonstrate how ecological connectivity, slow/mismatched management timescales (between environmental change and management actions), and different management actions can influence environmental quality in connected ecosystems.

A set of differential equations (see  Appendix 1) was used to represent a downstream ecosystem (e.g., a coastal fishery) whose quality is biophysically affected by an upstream ecosystem, such as nutrient loading in an agroecosystem (e.g., a situation like that described in Box 1). For the dynamics of the downstream ecosystem, we assumed a classic tipping point behavior of the environmental quality of greatest value to humans (i.e., an ES), where this quality can be downgraded by the loading of nutrients in an upstream system (e.g., Thrush et al., 2021). A social–ecological feedback is present, whereby degraded quality of the downstream ecosystem may lead to management actions (A) that limit runoff of nutrients from the upstream ecosystem (and therefore the concentrations of nutrients in the downstream ecosystem). These management actions could be triggered by environmental concern, decline in income, or loss of other benefits from the ES.

As expected, the ecological connectivity between upstream pollution and the downstream ecosystem can trigger transgression of a tipping point in the downstream ecosystem (Figure 4, black line). A sufficiently strong upstream feedback (A₀) on the “social” side of this social–ecological system can stabilize the downstream ecosystem against this collapse by creating actions that decrease nutrient levels in the downstream ecosystem (e.g., riparian planting and/or the restoration of wetlands; Figure 4, blue and red lines). Strong upstream feedbacks can prevent regime shifts and tipping points (no bistability in red line in Figure 4). That management or human behavioral feedbacks can shift the position of an ecological regime shift or remove it entirely has previously been demonstrated in theoretical (Lade et al., 2013) and empirical (Lade et al., 2015) social–ecological systems. The strength of this upstream feedback will depend on several factors such as sufficiently strong incentives, trust in management agencies, and public support.

The speed of management response also strongly influences the emergent social–ecological dynamics (Figure 5). Specifically, if the management response is strong but delayed, an oscillatory dynamic can result. In response to a disturbance, such as a sudden inflow of nutrients or change in policy affecting land clearing, management actions are likely to bounce between under- and overresponding with similar consequences for the quality of the ecosystem (Figure 5, red line). A faster management response (i.e., greater temporal matching of management to changes in ecosystem state) eliminates this oscillation (Figure 5, blue line). A moderate delay and a moderate management response (Figure 5, green line) could result in a permanent shift in ecosystem condition due to the system passing a tipping point prior to any management action.

Slow changes in pollutant loads, resulting in slow changes in environmental quality, challenge environmental management (Hughes et al., 2013). Over long time scales, environmental management often fails to respond to slow environmental changes even if the absolute change has been large. This is sometimes known as the “shifting baselines” syndrome (Soga and Gaston, 2018). Ecological time lags have hindered effective management in the Baltic Sea (see Box 1), for example, because the legacy of nutrient inputs over decadal timescales has saturated the marine ecosystem and “slow to recover” ecological processes that facilitate removal mediated by key species and regional biodiversity are the most affected.

The management interventions modeled here correspond to the reactive solutions that are often used in practice. In the model, once the intervention ceases, then pollutant inflows return to previous levels. Instead, future-response-focused management that aims at transformative changes, such as fundamental changes in agricultural systems or landscapes (e.g., riparian planting and buffer zones), should be a goal for long-term sustainability. We used this simple model to illustrate the importance of cross-domain connections, but it is important to note that it studies only a single stressor, whereas much of the complexity of environmental management stems from the presence of multiple stressors (Côté et al., 2016; Thrush et al., 2021; see Boxes 1–3).

While an ecosystem-based approach to management has been promoted by the research community for almost 20 years (Christensen et al., 1996), implementation has been slow and patchy. We argue that a focus on the social–ecological properties of cross-domain connections demands new management and research priorities, and this will assist effective implementation. These priorities include the need to recognize, in governance and in actions, the dynamic implications of slow decision making across linked domains and to enhance upstream social–ecological feedbacks, thereby reducing management response times across all ecological domains. Priorities include the following:

Current global trends in environmental degradation are sobering and necessitate a significant shift in our efforts to curb the damage. We argue that these 7 priorities can help to navigate research and management toward the situation depicted in Figure 6, where cross-domain connections and feedbacks are explicitly addressed in environmental management. This situation would lead to ecosystems that are more resilient to stress (blue and red lines in Figure 4) and management decisions that are effective at stabilizing and preserving ecosystem functions and services (blue line in Figure 5). Importantly, leveraging cross-domain linkages as early warning signals of future downstream change may alleviate temporal scale mismatches (lags) in management responses to change. The social subsystem may provide an important nexus for managing multidomain social–ecological systems, though this will require researchers and policy makers to step outside of traditional domain boundaries.

Aotearoa New Zealand’s research system has contributed to addressing these priorities through its National Science Challenges. These are large, collaborative, transdisciplinary projects involving extensive codevelopment. However, they were domain constrained by the funding agency, and this article is a product of a workshop to explore commonalities and connections across 3 of these challenges (“Biological Heritage” [Terrestrial], “Our Land and Water” [Freshwater], and “Sustainable Seas” [Marine]).

We thank Jasmine Low for assistance with the artwork. We also thank the editor and 2 reviewers for their thoughtful suggestions that improved the depth and interest of the manuscript.

The workshop was funded by the New Zealand National Science Challenges: Sustainable Seas (Tipping Points Project; CO1X1515; ST) and New Zealands's Biological Heritage (Project 3.1; 1516-44-004; JT), established by the Ministry of Business, Innovation and Enterprise, New Zealand. The writing of this manuscript was also funded by the Sustainable Seas National Science Challenge Project 1.1: Ecological responses to cumulative effects (C01X1901). RG-G was supported by the New Zealand Rutherford Foundation Postdoctoral Fellowship and the Walter and Andrée de Nottbeck Foundation during the writing of this manuscript.

No competing interests to declare.

Contributed to the workshop and conception of ideas: All authors.

Obtained funding for the workshop: JT, ST.

Organized the workshop: JT, ST, JY, JH, CP.

Contributed to the model development and results: VD, SL.

Wrote the first manuscript draft: RG-G, ST, JT.

Revised the first draft and approved final submission: All authors.

We assumed that the strength of management action (A₀) changes with time delay (τ) toward a target strength T(ES), where ES is the environmental quality of the downstream ecosystem/ES. We implemented these assumptions mathematically using a first-order autoregressive model,

We implemented the tipping point of environmental quality using Holling Type III reproduction with linear mortality and a linear effect of pollutant inflow r(A):

Here, h is the approximate position of the tipping point, p is the sharpness of the tipping point, a is the strength of the inflow’s impact on environmental quality, and we have normalized ES so that its maximum value is 1.

It remains to specify the linkages between the downstream environmental system and upstream management actions (A), r(A) and T(ES). For the effect of upstream management, we make the simple assumption that pollutant inflow responds linearly to management action, so that

where P is the pollutant inflow without management action. Any delays in response to management action can be incorporated to the delay parameter τ.

Control theory classifies the management response to changes of environmental quality into one of (or a combination of) the following types:

• Proportional control, where management action responds to the difference between the current environmental quality and a target environmental quality;

• Differential control, where management action responds to the rate of change of environmental quality (result of upper limit on timescales [generational]);

• Integral control, where management action responds to both the difference between current and target environmental quality and the time spent away from the target environmental quality.

Control theory discusses advantages and disadvantages of each of these types of control (Åström and Murray, 2012). Here, we assumed proportional control:

where E₀ is the target environmental quality, and A₀ sets the effectiveness of management action.

How to cite this article: Gladstone-Gallagher, RV, Tylianakis, JM, Yletyinen, J, Dakos, V, Douglas, EJ, Greenhalgh, S, Hewitt, JE, Hikuroa, D, Lade, SJ, Le Heron, R, Norkko, A, Perry, GLW, Pilditch, CA, Schiel, D, Siwicka, E, Warburton, H, Thrush, SF. 2022. Social-ecological connections across land, water, and sea demand a reprioritization of environmental management. Elementa: Science of the Anthropocene 10(1). DOI: https://doi.org/10.1525/elementa.2021.00075

Domain Editor-in-Chief: Alastair Iles, University of California Berkeley, Berkeley, CA, USA

Knowledge Domain: Sustainability Transitions