Identifying indicators

The planning and execution of this research involved a first phase of literature review aimed at finding indicators to determine the risks associated with DSM activities based on the current state of the DSM industry. The search focused on peer-reviewed articles, reports and online databases available on Google, which yielded approximately 70 articles, forming the bases for the analysis. The indicators were carefully chosen to capture both direct and indirect risks associated with DSM activities, encompassing immediate operational impacts as well as long-term consequences for various stakeholders and ecosystems. Those that were measurable or assessed with standard qualitative and quantitative methods in literature were considered, to ensure data validity. On the other hand, indicators from any unsupported evidence of literature and those with significant overlap or redundancy with other selected indicators were excluded.

The indicators were categorized into three major classes: i) environmental indicators: assessing the likely impacts of DSM on the environment; ii) economic indicators: evaluating cost factors in mining, including capital and operational expenditures, as well as environmental costs; and iii) social indicators: focusing on social aspects, including potential and conflicting interests of Indigenous communities. Environmental indicators were selected based on their ability to measure both immediate ecosystem disruption and long-term environmental consequences. Through this, we identified 8 environmental indicators, i.e., biodiversity risk; vulnerability to coastal states; pollution; troublesome sea life; knowledge gaps; greenwashing argument; ambiguous guidelines; and habitat risk. Detailed descriptions of each of these are provided in S1 Table in S1 File. Economic indicators were chosen to reflect both micro and macro-economic implications, ranging from direct operational costs and profitability risks to broader impacts on existing industries and national economies. We identified 12 economic indicators consisting of the following: private profit over public good; risk on commercial fisheries; ecological cost; technological cost; loss and profitability risks; tourism; royalty payment regimes; knowledge gap on EAF; terrestrial mining vs DSM; multiple shareholders; lack of special provision; and contractual violations. We provide detailed descriptions of each of these indicators in S2 Table in S1 File. Social indicators were selected to capture the complex interplay between DSM operations and societal well-being. Total 8 social indicators were identified, including: political implications; social stigma; food security and safety; lack of effective control; violation of the law; degraded reputation; livelihood and employment loss; and participatory rights. Detailed descriptions of each of these are provided in S3 Table in S1 File.

S4 diagram in S1 File illustrates the interconnected risks associated with DSM across environmental, economic, and social dimensions, highlighting the importance of a comprehensive assessment framework. Environmental indicators are crucial as they address the potential for irreversible damage to deep-sea ecosystems, emphasizing the need for conservation measures and adherence to global sustainability goals. Economic indicators are essential for evaluating the feasibility of DSM operations, particularly considering their implications for existing maritime industries and the economies of developing maritime nations. Equally important are social indicators, which provide insights into the broader social implications of DSM, including community well-being, governance, and employment. This integrated framework offers a structured approach to understanding and evaluating the multifaceted risks associated with DSM, enabling more informed and sustainable decision-making processes.

Data collection

After identifying the risk indicators, the questionnaires were developed for obtaining the responses from experts in the field. The questionnaire consisted of 2 sub questions for each individual major question: i) what is the likelihood that the proposed risk burden from DSM will affect the social-ecological systems; and ii) how significant would the resulting impact be? Interviews were conducted with nine experts, including representatives from NGOs, academia, and a DSM mining company, to capture perceptions of DSM risks. Detailed information is provided in (Table 1). Experts were asked to provide their views on the 41 identified questions, encompassing environmental, economic, and social risk indicators. The responses obtained were transcribed and consolidated into different categories using Microsoft excel.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Expert profile for risk indicators interview.

https://doi.org/10.1371/journal.pone.0320888.t001

We collected the indicators for this study through an extensive literature review. Expert opinions were sought solely to verify the relevance of the data, and their involvement was limited to providing feedback on pre-existing information. No personal or sensitive data were collected. Due to the non-sensitive nature of the data, written consent was not required. However, verbal consent was documented through meeting notes to ensure transparency and compliance with ethical standards.

Construction of Fuzzy Cognitive Map (FCM)

This study employs a Fuzzy Cognitive Map (FCM) to assess the perceived influence of DSM on marine ecosystem services. FCMs, originally proposed by Kosko [31], are computational models that combine fuzzy logic, neural networks, and cognitive mapping techniques [31,32]. These models are composed of several components and the causal connections between them, which together create a neural network [32,33]. Every component is given an activation value (AV), i.e., the degree of impact it has on the network of between 0 and 1, and each relationship is allocated a weight that represents the degree of causality, ranging from -1 to 1 (where 1 and -1 signify full positive and negative causality, respectively). Fuzzy logic is employed to transform both quantitative and qualitative data into a common format, which is then used to determine the AVs and weights [34]. In the present study, the FCM framework was applied to a hypothetical DSM project, integrating various risk indicators and expert opinions. Qualitative data from expert interviews have been used to calculate total risks multiplying the likelihood by the severity using the following equation,

(1)

Where:

Likelihood =  the probability of the risk event occurring (from 1 to 5).

Impact =  the severity of the consequences if the event occurs (from 1 to 5).

The risk score is used to determine the weight in the FCM, where the weight represents the degree of causality or influence between components. For example, a risk score of 25 corresponds to a weight of + 1, indicating a strong positive causality between the two components involved. This means that when both the likelihood of occurrence and the impact are high, the relationship between the components is considered fully positive, signifying that the risk has a significant influence on the network.

The second phase of the literature review is aimed at discovering the most significant indicators, which are linked to circular solutions (CS) encompassing circular economy, new technology and recycling. The phase involved reviewing 21 published sources (S4 Table in S1 File) to determine which keywords were most frequently associated with impacts on circular solutions. Each publication was categorized by the first author’s name and given a score of 1 when an indicator was mentioned and 0 when it was not. The assessment involved adding all of the scores and determining the average score per indicator across all variables using the method described by Cordova-Pozo and Rouwette [35]. The Indicator Score was calculated using the formula below,

(2)

The calculated Indicator Scores were subsequently used to determine the weights in the FCM, with each weight being directly proportional to the Indicator Score. For instance, an indicator with a score of 0.81 would have a corresponding weight of 0.81, reflecting the strength of the causality between components. These weights were then input into the “Mental Modeler” software to construct the FCM, which modeled the relationships between components based on their calculated weights, enabling a comprehensive analysis of the factors influencing circular solutions. The overall risk assessment methodological flowchart is presented in (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Framework for risk assessment of deep-sea mining using fuzzy cognitive mapping.

https://doi.org/10.1371/journal.pone.0320888.g001

The figure depicts the stepwise framework for assessing risks associated with DSM. The process involves three phases. Phase 1 begins with a literature review to identify risk indicators, categorized into environmental, economic, and social aspects. Phase 2 entails expert interviews to evaluate the likelihood of risks and the significance of impacts through a questionnaire. The responses are transcribed and consolidated for analysis. Phase 3 incorporates a review of published sources to analyze the impact of indicators on circular solutions using keyword analysis, scoring systems, and average score calculations. The outputs from all phases feed into the development of the FCM for a comprehensive risk assessment.

To quantify the FCM, we calculated the following metrics [36,37]:

• Number of Concepts (N): The total number of unique variables or nodes in the FCM;
• Number of Connections (C): The total number of directed edges or connections between the concepts;
• Connections per Concept (C/N): The average number of connections per concept;
• Density (D): An index of connectivity, calculated by dividing the number of connections by the maximum number of connections possible between N variables;
• Driver Variables: Positive outdegree and zero indegree, indicating they drive changes in the system;
• Receiver Variables: Positive indegree and zero outdegree, indicating they are influenced by other variables; and
• Ordinary Variables: Non-zero indegree and outdegree, indicating they both influence and are influenced by other variables.

The FCM model developed in this study (Fig 2) represents a system with 30 key elements interconnected through 41 relationships (Table 2). The density of the FCM 0.05, which is calculated as the ratio of the actual number of connections to the total possible connections in the map. On average, each component has approximately 1.36 connections. This metric provides insights into the average connectivity within the map. It has 2 driver components and 28 receiver components, indicating a dynamic system with multiple feedback loops and a moderate level of complexity (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. General Fuzzy Cognitive Map (FCM) statistics.

https://doi.org/10.1371/journal.pone.0320888.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Concept categorization and network metrics in the Fuzzy Cognitive Map (FCM).

https://doi.org/10.1371/journal.pone.0320888.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Fuzzy cognitive model of deep-sea mining and circular solutions for future metal supply.

The constructed Fuzzy Cognitive Map (FCM) represents a complex system with 30 key elements linked by 41 interconnections. It features 2 driver components, “deep-sea mining” and “circular solutions”, which significantly influences many other elements. “Deep-sea mining” emerges as the element with the highest centrality, highlighting its pivotal role and extensive connectivity within the system. Its extensive connectivity highlights the widespread social-ecological implications of DSM, emphasizing the importance of carefully evaluating mining practices and exploring circular economy strategies to mitigate negative consequences.

https://doi.org/10.1371/journal.pone.0320888.g002

In the context of FCM, the outdegree, indegree, and centrality of variables are key metrics used to evaluate the role and influence of each concept within the system. These measures help categorize variables as driver, receiver, or ordinary based on their connectivity and the direction of influence within the network. The outdegree of a variable refers to the cumulative strength of the connections that exit from that variable. It quantifies the influence that the variable exerts on other variables in the system. The outdegree is calculated as the sum of the weights of all outgoing edges from a variable. The indegree of a variable measures the cumulative strength of the connections entering that variable. It indicates how much a variable is influenced by other variables in the system. A higher outdegree indicates that the variable has a greater influence on other variables in the system. A higher indegree suggests that the variable is more influenced by other components in the system [36,37].

The centrality of a variable is the total sum of its outdegree and indegree, indicating its overall importance within the system. A variable with high centrality is more influential, either as a driver of change (outdegree) or as an important receiver of influence (indegree) and it is measured by the formula below [38],

(3)

In this analysis “deep-sea mining” has the highest centrality (24.28), indicating its significant influence and connectivity within the map (Table 3).

In the current study, the FCM approach was preferred over methods like the Analytic Hierarchy Process (AHP) due to its ability to model complex, dynamic systems with interdependencies and feedback loops [39]. Unlike AHP, which is static and focuses on hierarchical comparisons [40], FCM captures the non-linear interactions between ecological, economic, and technological factors [41]. FCM also integrates both qualitative and quantitative data, accommodates uncertainty, and offers a deeper understanding of causal relationships [42], making it more adaptable for evaluating the interconnected impacts of DSM on marine ecosystem services.

To address uncertainties inherent in the FCM approach, we conducted scenario analysis to examine the robustness of model outcomes under different assumptions. By systematically varying key connection weights and testing alternative configurations of the model, we assessed the extent to which uncertainties in expert-defined relationships influence system behavior. The results demonstrated that while some concepts exhibited sensitivity to specific changes, the overall system dynamics remained consistent across scenarios, reinforcing the model’s reliability.

The “With” and “Without” DSM analyses

Running the Mental Modeler in FCM involves assigning values ranging between -1 and + 1 to represent the influence or impact of one element on another within a model. Assigning different values in the model is used to simulate and predict the outcomes of different scenarios. When setting up scenarios in FCM, a weight of 0.1 might be assigned to a variable to reflect that this variable has a minor or limited influence over other variables within the FCM model. To the best of our knowledge, there is no existing research that establishes the current state and values of the social, environmental, and economic concepts related to DSM that could be incorporated into the creation of the FCM needed for our analysis. Therefore, following the concept of Kokkinos et al. [32] and Hatziioannou & Kokkinos [43], we consider the steady state of the FCM model as the status quo scenario.

The effects of deep-sea mining

The findings of the “With” DSM analysis for all the studied indicators are presented in (Fig 3). The immediate observation from the figure is that DSM increases the risks associated with all indicators. The highest likely risk imposed by DSM is on coastal states as measured by the “vulnerability to coastal states” indicator at an increase of 13% relative to the “Without” DSM. for example, mining activities in the CCZ pose a significant threat to Small Island Developing States in the region, compounding the existing challenges these nations face due to the continuous rise in sea levels [45]. These dual threats, environmental degradation from DSM and the long-term impacts of climate change, create a precarious situation for these small island countries, with profound implications for their ecosystems, economies, and societies.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Relative changes in DSM impacts across environmental, social and economic indicators.

This figure illustrates the percentage changes from a steady state (our assumed “Without” DSM scenario) in activation values - a measure of the impacts of DSM across various indicators. The indicators, listed on the y-axis, reflect different dimensions of environmental, social, and economic impacts. The x-axis represents the percentage change in activation values for each indicator. Activation values are critical in this context as they help quantify the magnitude of DSM’s impact, providing a clearer picture of how various elements of the system are affected. Percentage changes in these values highlight key insights, such as the increased vulnerability of coastal states and reputational risks, which emerge as the most significantly impacted indicators.

https://doi.org/10.1371/journal.pone.0320888.g003

Social indicators that register high increases in risks are “violation of law,” “participatory rights,” “lack of effective control,” and “degraded reputation,” with relative increases of 8% to 11%. The “violation of law” as a social risk in DSM highlights legal concerns stemming from insufficient environmental assessments. For instance, in Norway, advancing DSM without robust scientific evidence could breach national and international laws [46]. Such actions pose a risk of undermining environmental protections and may provoke legal challenges, emphasizing the necessity of adhering to regulations in order to preserve regulatory integrity and public trust [47]. “Participatory rights” in the context of DSM refer to the procedural rights of stakeholders, including Indigenous peoples, local communities, and the general public, to actively engage in decision-making processes. Currently, the legal framework governing DSM often falls short of fulfilling state obligations to ensure meaningful stakeholder involvement. Incorporating indigenous peoples into DSM management is not just a procedural requirement but a fundamental aspect of their comprehensive, tradition-based connection to the environment, as acknowledged by international initiatives and the regulations of Pacific States [48,49]. Similarly, “lack of effective control” in DSM refers to the insufficient oversight and authority of sponsoring states over the companies they endorse for mining activities. Furthermore, the ISA has not clearly defined what constitutes “effective control,” leading to ambiguities. For example, in the case of TOML, Tonga’s ability to fulfill its regulatory responsibilities is uncertain, highlighting the critical need for clearer definitions and stronger governance mechanisms [50]. On the other hand, “degraded reputation” highlights the damage to a state’s image due to controversial decisions. In particular, Norway’s decision to open its seabed for commercial DSM, despite warnings from scientists and environmental experts, has drawn criticism potentially eroding trust and credibility in its environmental commitments [51].

We also see large economic risks, especially in areas such as “contractual violation,” “lack of special provision,” “multiple shareholders,” “knowledge gap on EAF,” “loss and profitability risks”, “private profit over public good” and “unacceptable royalty payment regimes”. “Contractual violation” in DSM refers to the potential breach of agreements between companies and regulatory bodies, such as the ISA. For example, if TOML, a subsidiary sponsored by Tonga, fails to meet its contractual obligations, the sponsoring state (Tonga) could face significant financial liabilities [52]. This adds to the country’s burden, as the environmental damage from DSM activities could far exceed any economic benefits. Besides, the “lack of special provision” in DSM highlights the absence of differentiated responsibilities under the ISA framework. Both developing and developed countries sponsoring mining activities are held equally accountable for environmental damage caused to the ocean floor. For developing countries this creates significant challenges, as they lack the financial and technical resources of developed nations to manage and mitigate such damages [53].

The uniform application of rules creates disproportionate challenges for less-resourced states, highlighting the need for equitable provisions that consider the varying capacities of sponsoring countries. “Multiple shareholders” in DSM refers to the complexity and challenges arising from the involvement of various entities, including foreign companies and international investors, in the ownership and control of a mining operation [50]. Although a company may seem to be domestically owned, its management and financial control are often in the hands of foreign entities or shareholders. This complicates the sponsoring state’s ability to hold foreign stakeholders accountable for legal, financial, or environmental issues, increasing the risks linked to multiple levels of oversight and jurisdiction. The “knowledge gap on EAF (Economic Assistance Fund)” refers to the uncertainty surrounding the distribution of shares from DSM as compensation to terrestrial mining countries. Limited research and unresolved mechanisms for payment calculations leave the extent of compensation unclear, emphasizing the need for further study and transparent frameworks [54]. Furthermore, “loss and profitability risks” in DSM encompass several challenges, including: risk of loss and liability, where sponsoring states bear responsibility for environmental damage and must adhere to stringent international standards, potentially burdening taxpayers loss of investment [50,53], as seen in cases like PNG’s $120 million stake in Nautilus Minerals, which faced bankruptcy [55]; risk of reduced benefit, where financial and social promises of DSM often fail to materialize due to project failures and public rejection [48,56,57]; and lower success rate/profitability, as DSM lacks cost advantages over land-based mining and poses greater environmental risks, casting doubt on its long-term viability [58,59]. In DSM, the indicator “private profit over public good” reflects concerns over company monopoly and business-sided profits.

Companies with market capitalizations far exceeding the GDP of many small nations dominate the sector, leveraging global resources for legal and negotiation advantages, fostering monopolies despite the seafloor being the “common heritage of mankind” [60,61]. Moreover, cost-benefit analyses, like those conducted by MIT, reveal that miners earn nearly three times the profits compared to the public benefits distributed by the ISA [62]. Critics argue that this dynamic enriches a small elite while failing to deliver prosperity to indigenous communities or society at large [63]. Similarly, the “unacceptable royalty payment regimes” highlights concerns over inequitable compensation structures. For example, the African Group opposes time-varying royalty rates (e.g., 2%/6%) and price-varying royalties (e.g., 2%/5% to 9%), arguing that these regimes fail to provide fair compensation to humanity, do not maximize ISA revenues, and effectively subsidize DSM at the expense of land-based mining [64]. These payment structures are criticized for prioritizing industry profits over equitable benefit-sharing and global interests. Collectively, the challenges reflect the intricate economic and governance issues that may arise from engaging in DSM, where various stakeholders could encounter conflicts, regulatory ambiguities, and financial disagreements. Our results demonstrated moderate risks associated with ecological and technological costs, greenwashing arguments, political implications, comparisons between terrestrial and DSM, and tourism indicators. The findings suggest that, while these sectors are less immediately affected, they still raise significant problems that require attention.

In contrast, the scenario analysis identifies minor risks associated with a number of cross-border variables, including “social stigma,” “livelihood and employment loss,” “commercial fisheries,” “biodiversity risks,” “pollution,” “food security and safety,” and “marine habitat.” The smaller risks calculated for these indicators shows that despite the substantial challenges posed by engaging in DSM, certain aspects of environmental and social stability may remain relatively resilient to DSM impacts as a result of the assumed significant contribution of CS.

DSM and ecosystem services

The benefits that human societies obtain from ecosystems, known as ecosystem services, are essential to the well-being of both human populations and the environment. These services are commonly classified into four main categories: provisioning, supporting, regulating, and cultural services [65]. Every category represents a unique aspect of the ecosystem’s functionality, which contributes to the overall well-being and long-term viability of our planet. DSM presents serious challenges to these capabilities, specifically in the fragile and usually uncharted deep-sea ecosystems. (Fig 4) depicts the effects of DSM on distinct ecosystem services, emphasizing the related risks and benefits. The figure shows that DSM elevates risk across several ecosystem services, including a 2%, 3% and 5% increases in risks to commercial fisheries, pollution and biodiversity; to marine habitats; and to tourism, respectively. DSM threatens commercial fisheries through conflicts with fishing industries, especially in regions like the CCZ, where discharge water, heavy metal plumes, and mining noise can harm marine species and disrupt tuna fisheries [6,66]. Pollution from DSM can travel vast distances, with discharge plumes potentially reaching areas like Hawaiian and Kiribati waters, causing widespread environmental degradation [67]. Biodiversity is at risk as DSM activities may cause permanent damage to pristine habitats, leading to species loss and ecosystem fragmentation, particularly in biodiversity hotspots and deep-sea ecosystems [68,69]. Mining also disrupts marine habitats, with techniques like sediment removal and noise pollution threatening species such as the gastropod Dracogyra subfusca [70]. Lastly, tourism in affected regions may suffer due to the loss of unique marine species and habitats, diminishing the aesthetic and recreational value of these areas and leading to reduced tourist interest and revenue [71]. Overall, the figure suggests that engaging in DSM would likely pose significant risks to ecosystem services and human well-being.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Impact of deep-sea mining (DSM) on ecosystem services and human well-being.

The figure illustrates the pathways through which DSM increases risks to various ecosystem services and, in turn, affects human well-being. DSM is shown to increase risks to commercial fisheries, biodiversity, pollution levels, marine habitats, and tourism. Arrows represent the flow of these risks, moving from the scenario (DSM) to ecosystem services and subsequently impacting human well-being factors such as food security, safety, livelihood, and employment. The figure emphasizes the interconnectedness of environmental and human systems, illustrating those negative impacts on ecosystem services cascade into social and economic vulnerabilities. The percentage risk increments visually highlight the magnitude of potential impacts within each category.

https://doi.org/10.1371/journal.pone.0320888.g004