Understand systemic risk from mangrove ecosystem through network analysis.

Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Fundam Res . 2025 Nov 17;6(2):636–646. doi: 10.1016/j.fmre.2025.11.003 Search in PMC Search in PubMed View in NLM Catalog Add to search Understand systemic risk from mangrove ecosystem through network analysis Mimi Gong Mimi Gong a Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University, Beijing 100875, China b Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China Find articles by Mimi Gong a, b , Ke Yu Ke Yu c Center for Energy & Environmental Policy Research, Beijing Institute of Technology, Beijing 100081, China d School of Management, Beijing Institute of Technology, Beijing 100081, China Find articles by Ke Yu c, d , Qiang Huang Qiang Huang c Center for Energy & Environmental Policy Research, Beijing Institute of Technology, Beijing 100081, China d School of Management, Beijing Institute of Technology, Beijing 100081, China Find articles by Qiang Huang c, d, ⁎ , Yinglan A Yinglan A a Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University, Beijing 100875, China b Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China Find articles by Yinglan A a, b , Miriam Aczel Miriam Aczel e Institute for Water, Environment and Health, United Nations University, Richmond Hill, L4B 3P4 Ontario, Canada Find articles by Miriam Aczel e , Ye Li Ye Li c Center for Energy & Environmental Policy Research, Beijing Institute of Technology, Beijing 100081, China d School of Management, Beijing Institute of Technology, Beijing 100081, China Find articles by Ye Li c, d , Maofang Ran Maofang Ran a Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University, Beijing 100875, China b Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China Find articles by Maofang Ran a, b , Yan Cheng Yan Cheng a Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University, Beijing 100875, China b Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China Find articles by Yan Cheng a, b , Kaiji Li Kaiji Li a Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University, Beijing 100875, China b Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China Find articles by Kaiji Li a, b , Shen Qu Shen Qu c Center for Energy & Environmental Policy Research, Beijing Institute of Technology, Beijing 100081, China d School of Management, Beijing Institute of Technology, Beijing 100081, China Find articles by Shen Qu c, d, ⁎ Author information Article notes Copyright and License information a Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University, Beijing 100875, China b Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China c Center for Energy & Environmental Policy Research, Beijing Institute of Technology, Beijing 100081, China d School of Management, Beijing Institute of Technology, Beijing 100081, China e Institute for Water, Environment and Health, United Nations University, Richmond Hill, L4B 3P4 Ontario, Canada ⁎ Corresponding authors. [email protected] [email protected] Received 2025 Jun 24; Revised 2025 Sep 25; Accepted 2025 Nov 2; Collection date 2026 Mar. © 2025 The Authors. Publishing Services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). PMC Copyright notice PMCID: PMC13069846  PMID: 41971831 Abstract Mangrove deforestation amplifies systemic risks by worsening extreme weather events, impeding socio-economic development, and exposing governance vulnerabilities. Yet, the extent to which mangrove dynamics—both loss and restoration—interact with climate, socio-economic, and governance systems to mitigate systemic risk remains underexplored. Drawing on the economic concept of “product space,” we construct a Mangrove Multisystemic Risk Space, a network-based framework linking indicators across mangrove change, climate impacts, socio-economic development, and policy interventions. The network reveals a bipartite structure, with distinct clusters for mangrove loss and expansion, each surrounded by synergistic indicators. The mangrove loss cluster is tightly coupled with greenhouse gas emissions and climate extremes, while the expansion cluster aligns with renewable energy, economic growth, and population dynamics. Within this space, we identify two types of structurally significant indicators: “influential” (e.g., Ramsar site coverage) with high cascading potential, and “complex” indicators that require coordinated improvements across multiple dimensions, highlighting their systemic vulnerability. At the national level, the United States leads in achieving complex goals such as reducing extreme events, whereas New Zealand and Panama emerge as hubs of influential, well-performing indicators. These findings underscore the differentiated roles of mangrove-rich nations in mitigating systemic risk and call for strengthened global cooperation in mangrove conservation. Keywords: Mangrove ecosystems, Systemic risk, Cascading effects, Complex network analysis, Multidimensional evaluation Graphic abstract Open in a new tab 1. Introduction The rapid growth of the global economy has amplified interconnections across nations and regions [ [1] , [2] , [3] ], which triggers environmental systemic risk. Environmental systemic risk is defined as the complex and cascading risk that arises when ecological issues interact with climate and socio-economic systems, causing impacts exceeding single systems and difficult to reverse [ [4] , [5] , [6] , [7] , [8] ]. Specifically, within the context of mangrove ecosystems, we define mangrove systemic risk as the interconnected climate, socio-economic and governance vulnerabilities triggered by mangrove loss. Mangrove, as a unique ecosystem in coastal areas, plays a critical ecological and climatic role [ [9] , [10] , [11] ]. However, a total of 11,700 square kilometers has been lost globally since 1996 [ 12 ], with more than half of mangrove ecosystems facing collapse risk [ 13 ]. Such losses not only cause localized ecological damage but also trigger systemic environmental risks that cascade across climate and socio-economic systems. On the one hand, it reduces the ecosystem’s carbon sequestration capacity, while releasing large amounts of stored carbon through burning, worsening the global greenhouse effect [ 14 , 15 ]. The extreme climate events, such as tropical storms, hurricanes, and severe droughts, will in turn damage the mangrove habitat. On the other hand, mangrove ecosystems support nearly 800 billion juvenile fish, shrimp, crabs, and mollusks annually. Their degradation weakens ecological regulation capacity, resulting in damage to coastal fisheries and affecting the livelihoods of fishermen [ 16 , 17 ]. Additionally, the loss of ecological value and coastal protection capacity reduces eco-tourism income in coastal areas and increases vulnerability to erosion and storms, forcing governments to replace mangrove conservation funds with post-disaster reconstruction [ [18] , [19] , [20] ]. Therefore, the climate change and socio-economic issues triggered by mangrove degradation interact and form a vicious cycle, which seriously disrupts ecosystem balance and increases the vulnerability of socio-economic development. Existing research primarily analyzes mangrove loss from a single perspective—focusing either on natural or anthropogenic factors, which lacks a comprehensive analysis of the coupled effects of climatic, socio-economic, and other cross-system factors [ [21] , [22] , [23] , [24] ]. From the systemic risk management perspective, environmental control policies directly benefit ecosystems, yielding positive environmental outcomes. However, these policies also influence climate and socio-economic development through their impacts on ecosystems, leading to complex cascading effects and making it difficult to assess their overall effects [ 25 , 26 ]. Traditional environmental risk assessment method, such as life cycle assessment, traces pollutant impacts on ecosystems through material flow analysis [ 27 ]. Methods such as the Pressure-State-Response (PSR) model and system dynamics rely on linear causal assumptions and analyze environmental risks using a top-down approach [ 28 , 29 ]. In recent years, more integrated approaches, such as participatory system dynamics, have been developed to comprehensively capture the complex interrelationships among social, environmental, economic systems, and urban water systems, improving flood risk management [ 30 ]. Additionally, Bayesian hierarchical models have been used to analyze the impact of national regulatory strength on mangrove loss [ 31 ]. However, these methods have inherent limitations: they either rely on strong linear assumptions that conflict with real-world conditions, fail to capture driving factors from multiple systems, or cannot depict the effects of interactions among systems, thus failing to assess the systemic cascading risks of mangrove loss accurately. Complex network methods can effectively assess environmental systemic changes and their cascading effects by mapping the interaction and feedback mechanisms between different components [ 32 ]. However, the stability of the complex network established based on the correlation network is relatively unreliable [ 33 ]. The “product space” theory which initially developed in economic complexity research, constructs a network among products based on countries’ revealed comparative advantage in international trade, thus revealing opportunities for structural transformation and economic development [ 34 ]. This method has been widely used beyond national trade to technology innovation, resource efficiency, and regional development [ [35] , [36] , [37] ]. Moreover, it has been applied to environmental field and can be a good approach to understand systemic risk for effective environmental policies for the following reasons. First, the “product space” can assess the development of the entire system through quantifying the interactions among indicators. For example, in 2025, Ma et al. systematically captured this interplay by constructing an “SDG space,” which visualizes the disparities and development trajectories across countries in achieving Sustainable Development Goals (SDGs). They measured the interactions among different indicators, offering a comprehensive perspective on systemic progress [ 38 ]. Second, the concept of “comparative advantage” of the theory can be extended to grab the deeper connection of mutual promotion among indicators, where synergistic policies or technologies often share common knowledge, infrastructure, or institutional support, thus improving one indicator can drive the development of related indicators. Therefore, the conclusions drawn by this method offer significant policy guidance. In 2022, Penny et al. adopted a green-product network to identify low carbon transition pathways that combine lower CO₂ emissions with stringent environmental policies [ 39 ]. Inspired by these former research, Gong et al. used network techniques to study policy co-evolution between mangrove conservation and sustainable development [ 40 , 41 ]. Moreover, it has been proved that the network constructed by product space is more stable than correlation network, thus facilitating the long-term monitoring and update of policies [ 42 ]. These applications demonstrate the utility of the product space method in capturing latent synergies and dependencies in socio-environmental systems, making it a promising tool for identifying leverage points to mitigate environmental systemic risks. We extend the theory to mangrove systemic risk assessment by treating mangroves, climate, socio-economic factors, and governance indicators as “products” and constructing a four-dimensional collaborative network. Moreover, this study develops an integrated approach that combines betweenness centrality, eigenvector centrality, and complexity analysis to identify structurally important indicators within the network for targeted policy interventions. By aligning these network-based diagnostics with policy goals, decision-makers can prioritize the most critical indicators to reduce environmental systemic risk and advance Sustainable Development Goals (SDGs), such as SDG 8 (Decent Work and Economic Growth), SDG 13 (Climate Action), and SDG 14 (Life Below Water). To that end, this study aims to address four key scientific questions: (1) How can a systemic risk network be constructed to capture interdependencies among mangrove-related, climatic, socio-economic, and policy indicators? (2) Which structural and functional characteristics of the network reveal the mechanisms underlying the propagation of systemic risk across indicators? (3) How can a multidimensional evaluation framework identify key indicators for targeted policy intervention based on their systemic importance and complexity? (4) How can interdisciplinary strategies be tailored to specific regions to reduce mangrove-related systemic risk and promote integrated ecological, economic, and social development? To answer these questions, we first construct the Mangrove Multisystemic Risk Space. This network framework reveals the synergistic linkages between mangrove dynamics, climate change, socio-economic development, and governance indicators. This framework illustrates the cascading impacts of mangrove loss and identifies high-leverage indicators with strong cross-dimensional influence. We then develop a multi-criteria evaluation approach to pinpoint key indicators essential for sustaining long-term regional development and mitigating systemic risk. Finally, we demonstrate the framework’s policy relevance through case studies, such as Australia, offering tailored recommendations for enhancing systemic resilience and promoting sustainable integration. 2. Method 2.1. Synergistic relations between indicators By introducing the concept of product space from economics, this study calculates the synergy between four categories of indicators: 1) Mangrove indicators include mangrove loss caused by erosion and extreme climate events, human-induced mangrove loss, loss triggered by international trade, mangrove area growth, net mangrove area loss, total mangrove area, forest area, and forest carbon storage, etc. 2) Climate indicators include national targets for total greenhouse gas emissions and net emissions, the frequency of extreme climate events, sea level rise, and average temperature changes, etc. 3) Socio-economic indicators involve GDP, the proportion of industrial GDP, fixed asset investment, foreign direct investment, population size, non-renewable energy generation, and renewable energy generation, etc. 4) Government governance indicators include government efficiency, corruption levels, nationally determined contributions for emission reductions, ecological environment index, the number and area of Ramsar wetlands, and the capacity of marine protection personnel, etc. To reflect the negative impact of specific indicators on sustainable development, indicators representing mangrove loss, greenhouse gas emissions, and extreme climate events were treated negatively. We collected the above indicators from 60 countries with mangrove coverage, including Indonesia, Brazil, Australia and Mexico, etc. The data used in this study were obtained from several authoritative sources, including the International Monetary Fund’s Climate Change Indicators Dashboard (IMF-CID), World Bank Data, the United Nations Population Division, the International Monetary Fund (IMF), and OECD National Accounts. Mangrove loss attributed to final consumption embodied in international trade (M-4) is the sum of domestic consumption-related loss (M-5), consumption-related loss in neighboring countries (M-6), and consumption-related loss in distant countries (M-7) [ 43 ]. To maintain data consistency and stability, we represented each indicator with the average value from 2010 to 2016. Meanwhile, to eliminate extreme value fluctuations, we truncated and standardized the data. Detailed information on the categorization of indicators is provided in Table S3. In the network framework, we construct a systemic risk space where each indicator is assigned a score ranging from 0 to 100. A score closer to 0 indicates poorer performance and thus higher risk, while a score closer to 100 signifies better performance and lower risk. In parallel, the proximity values between indicators, ranging from 0 to 1, capture the strength of their mutual interaction. Higher proximity values suggest stronger synergistic effects, meaning that improvement in one indicator can facilitate improvement in another, thereby contributing to the reduction of systemic risk. In this case, we aim to analyze the mutual promotion patterns among indicators that can effectively promote the reduction of systemic risks and the advancement of SDGs using this network-based approach from a mathematical perspective, proximity is used to quantify the comparative advantage of a region in pursuing various goals. It can be succinctly expressed as the conditional probability that another objective is effectively achieved when one objective is attained at a high level within a specific region. This proximity is denoted by θ ( g, j ) and was calculated using Eq. 1 below: θ ( g , j ) = min { P ( R C A ( i , g ) > 1 | R C A ( i , j ) > 1 ) , P ( R C A ( i , j ) > 1 | R C A ( i , g ) > 1 ) } = ∑ i I ( R C A ( i , g ) > 1 ) | I ( R C A ( i , j ) > 1 ) max { ∑ i I ( R C A ( i , g ) > 1 ) , ∑ i I ( R C A ( i , j ) > 1 ) } (1) Here, g and j denote different specific mangrove, climate, socio-economic, or governance indicators. When the specified condition is met, I (·) = 1; otherwise, I (·) = 0. RCA ( i, g ) represents the comparative advantage of country i in achieving indicator g and it was calculated based on Eq. 2 below. R C A ( i , g ) = x ( i , g ) ∑ g ′ x ( i , g ′ ) ∑ i ′ x ( i ′ , g ) ∑ i ′ , g ′ x ( i ′ , g ′ ) (2) Here, x(i, g) represents the development level of region i for indicator g , and the symbol ∑ denotes the summation over all indicators. Specifically, ∑ x ( i, g’ ) represents the total development level of region i across all indicators, while the summation ∑ x ( i’, g ) represents the total development level of all regions for indicator g . When RCA ( i, g ) > 1, it indicates that region i has a significant comparative advantage over other regions for indicator g . Through a series of analytical calculations, we constructed a 64 × 64 Complex Multisystem Space network matrix that integrates mangrove, climate, socio-economic, and governance indicators. Each cell in the matrix represents the relationship between a pair of distinct indicators, effectively capturing their synergistic effects. The average proximity value is used as the cutting point, and relationships between indicators below this threshold are excluded when constructing the network space. 2.2. Mangrove multisystemic risk space 2.2.1. Modularity classification We employ a community detection algorithm to identify the optimal cluster partition by maximizing the modularity of the network. Specifically, the optimal number of clusters is selected based on the highest modularity score of the network space. Modularity is defined as the difference between the observed and expected number of intra-community edges, where the latter is based on a random network model [ 44 ]. Modularity values range from −1 to 1, with higher values indicating a more pronounced community structure, meaning that nodes within the same cluster are more densely connected to each other than to nodes in other clusters. This implies that indicators belonging to the same cluster exhibit stronger mutual relationships and function in a more coordinated manner. The specific calculation is provided in Eq. (3) below. Q = 1 2 m ∑ i j | A i j − k i k j 2 m | δ ( c i , c j ) (3) Here, A ij represents the connection weight between nodes i and j, k i and k j represent the degrees of nodes i and j , and m represents the total number of edges in the network. Additionally, when i and j belong to the same community or cluster, the value is 1; otherwise, it is 0. This study uses the Louvain algorithm to compute the optimal modularity. The algorithm consists of two steps: in Step 1, each node is moved to its neighbor’s community, the change in modularity is calculated, and the move that maximizes the modularity gain is chosen; in Step 2, nodes within the same community are merged into a super-node, the network is reconstructed, and the process is repeated until modularity no longer increases [ 45 ]. Compared to other algorithms, the Louvain algorithm is an unsupervised heuristic algorithm that can quickly optimize modularity and achieve optimal results in a short time, making it more suitable for large-scale networks with high efficiency requirements. 2.2.2. Betweenness centrality The internal and external structures of clusters in the network space are characterized by calculating the betweenness centrality of each indicator. Betweenness centrality captures a node’s influence by measuring the proportion of shortest paths between all pairs of nodes that pass through it. Nodes with high betweenness centrality act as critical “bridges” that facilitate communication and interaction across different clusters. The betweenness centrality was calculated using Eq. (4) below. C B ( v ) = ∑ s ≠ v ≠ t ∈ V σ s t ( v ) σ s t (4) Here, V is the set of all nodes in the network, σ s t is the number of shortest paths from node s to node t , and σ s t ( v ) is the number of shortest paths from node s to node t passing through node v . We use the Ulrik Brandes algorithm to compute the betweenness centrality of each node. This method optimizes the computation process, significantly reducing the time complexity [ 46 ]. Nodes with high betweenness centrality typically lie at the interfaces between distinct clusters in the network space. By identifying and enhancing these high-betweenness indicators, we can effectively strengthen the linkages between different clusters, thereby promoting integration and coordination across multiple subsystems. Improving such pivotal indicators can propagate positive influence throughout the network, simultaneously mitigating systemic risks, where interventions at these “bridge” nodes can cause amplified and network-wide consequences, ultimately supporting more resilient and sustainable development. We use Gephi 0.10.0 to visualize the Mangrove Multisystemic Risk Space network, employing the Force Atlas and Fruchterman-Reingold layouts to arrange node distribution. These algorithms iteratively adjust node positions until the system reaches a dynamic equilibrium [ 47 ]. In the visualization results, nodes represent mangrove, climate, socio-economic, and governance indicators, while edges denote the proximity values between them. Additionally, nodes with higher betweenness centrality appear larger in size and are more densely connected, typically occupying central positions within the network. In contrast, nodes with lower centrality tend to be smaller and are positioned near the periphery. 2.3. Multidimensional evaluation framework 2.3.1. Eigenvector centrality Eigenvector centrality measures the influence of an indicator within the network. An indicator attains a high eigenvector centrality if its development promotes the growth of other indicators, particularly when those affected indicators further stimulate additional developments. In this way, the indicator exerts a broad systemic impact across the network. After obtaining the proximity matrix from the Space model, the eigenvector centrality of each node was calculated using Eq. (5) . x i = 1 λ ∑ j w i j x j (5) Specifically, a node’s eigenvector centrality equals the sum of the eigenvector centralities of its neighboring nodes, each weighted by the corresponding proximity, and divided by the largest eigenvalue of the network. Here, the proximity values from the upper triangular portion of the 64 × 64 matrix were used as edge weights w ij (only considering the upper triangular part to avoid redundant calculations). x i and x j mean the eigenvector centrality of indicator i and j , while λ is the largest eigenvalue. The power iteration algorithm is used to approximate the principal eigenvector of the node adjacency matrix iteratively. Based on the eigenvector centrality of indicators, we calculated the eigenvector centrality of various regions using Eq. (6) below. r e g i o n _ c e n t r a l i t y i = 1 K ∑ j = 1 K ( D a t a [ i , j ] 100 × i n d i c a t o r _ c e n t r a l i t y j ) (6) Here, Data [ i,j ] is the scores of indicator j in region i , while indicator_centrality j is the eigenvector betweenness of indicator j . We computed the regional eigenvector centrality by taking a weighted average of the eigenvector centralities of all 64 indicators, using the corresponding indicator scores as weights. Since the indicator scores range from 0 to 100, each score was normalized by dividing it by 100 before applying it as a weight. 2.3.2. Complexity Complexity measures an indicator’s structural importance within the network. Specifically, if an indicator tends to have higher values in regions where other indicators also score highly—i.e., regions with more comprehensive development—this suggests that its progress often relies on the advancement of other indicators. Such indicators are described to have high complexity, as they reflect and depend on broader systemic development. For matrix Data with N regions and K indicators, we initially assume that the complexity value for each indicator j , denoted as origin_indicator_complexity j , is set to 1. Meanwhile, origin_region_complexity i represents the initial complexity of region i , which equals the average value of region i across the K indicators, as shown in Eqs. (7) , (8) . o r i g i n _ i n d i c a t o r _ c o m p l e x i t y j = 1 ∀ j (7) o r i g i n _ r e g i o n _ c o m p l e x i t y i = 1 K ∑ j = 1 K D a t a [ i , j ] (8) However, complexity values at both the indicator and regional levels are dynamic and updated through iterative calculations. Specifically, for indicator j , the scores of the N regions are multiplied by their regional complexity and then divided by the sum to obtain the indicator complexity, denoted as indicator_complexity j . In other words, the complexity of each indicator is the weighted average of the complexity across all regions, with the weights being the scores of each region. This process is described by Eq. (9) below. i n d i c a t o r _ c o m p l e x i t y j = ∑ i = 1 N D a t a [ i , j ] × r e g i o n _ c o m p l e x i t y i ∑ i = 1 N D a t a [ i , j ] (9) For region i , the scores of the K indicators are multiplied by their respective complexity and then divided by the sum to obtain the regional complexity, denoted as region_complexity i . In other words, the complexity of each region is the weighted average of the performance across all indicators, with the weights being the complexities of each indicator. Therefore, the regional complexity was calculated based on Eq. (10) below. r e g i o n _ c o m p l e x i t y i = ∑ j = 1 K D a t a [ i , j ] × i n d i c a t o r _ c o m p l e x i t y j ∑ j = 1 K i n d i c a t o r _ c o m p l e x i t y j (10) To ensure the effectiveness and rigor of the iterative process, the complexity at both the indicator and regional levels must be calculated and compared with the previous iteration’s results after each iteration. Only when the complexity changes for all regions and indicators are smaller than the set tolerance will it indicate that dynamic equilibrium has been reached, allowing the loop to terminate. Proximity measures the synergistic relationship among indicators. Betweenness centrality highlights key “bridge” indicators that facilitate co-development across domains. Eigenvector centrality identifies indicators with broad systemic influence, whose improvement can generate cascading positive effects throughout the network, thus accelerating sustainable development. Complexity analysis, adapted from economic complexity metrics, reflects the inherent difficulty of achieving a given indicator based on its dependence on the fulfillment of multiple simpler indicators—signifying greater vulnerability to systemic disruptions. For instance, achieving gender parity in earnings often presupposes broader economic gains in labor income. Table 1 lists the concept differences between different evaluation approaches and their definitions. Table 1. Concept clarification of the complex multisystem space . Concept Object Clarification Proximity Edge It reflects the complementary relationship between indicators. The larger the proximity, the stronger the synergy between indicators, meaning that the development of one indicator helps drive the growth of another. Betweenness centrality Node Betweenness centrality captures the extent to which a node lies on the shortest paths between other nodes, highlighting key “bridge” indicators that facilitate co-development across domains. Eigenvector centrality Node Eigenvector centrality identifies indicators with broad systemic influence, whose improvement can generate cascading positive effects throughout the network, thus accelerating sustainable development. Eigenvector centrality provides a measure of the impact of a node in a connected network. Complexity Node Complexity analysis, adapted from economic complexity metrics, reflects the inherent difficulty of achieving a given indicator, whose achievement depends on the fulfillment of multiple simpler indicators, signifying greater vulnerability to systemic disruptions. For instance, achieving gender parity in earnings often presupposes broader economic gains in labor income. Open in a new tab 3. Results Using the unsupervised Louvain algorithm, a clustering analysis was performed on 64 indicators, which were ultimately divided into three clusters. These clusters are represented in the network Space by light yellow, green, and purple, respectively. The first cluster includes indicators related to mangrove loss, including M-1 (mangrove loss caused by erosion) to M-9 (mangrove gross loss), and is classified as the Mangrove Loss Cluster. The second cluster includes indicators related to mangrove increase and coverage, from M-10 (mangrove gross gain) to M-13 (mangrove area), and is classified as the Mangrove Expansion Cluster. The third cluster includes indicators related to government measures and government efficiency, classified as the Government Measures Cluster. The modularity of the clusters computed by the algorithm is 0.351, indicating, based on experience, that the clustering results are highly reliable. To better visualize the results, a threshold of 0.37 for edge proximity was set, and 899 connecting edges between nodes were plotted in the network space. Additionally, edges with proximity greater than 0.8 were highlighted to emphasize highly correlated node relationships ( Fig. 1 ). Fig. 1. Open in a new tab Mangrove multisystemic risk space. Background colors indicate different clusters identified via community detection. The color of the nodes represents different types of indicators based on prior knowledge, with labels indicating each type (M for mangrove indicators, S for socio-economic indicators, G for governance indicators, and C for climate indicators). The size of the nodes represents the betweenness centrality of the indicators (larger nodes correspond to higher betweenness centrality). The thickness of the edges corresponds to the proximity between indicators (thicker edges indicate higher proximity). 3.1. Mangrove loss exhibits a strong synergistic relationship with climate change Cluster 1 (Mangrove Loss) contains more indicators overall than Cluster 2 (Mangrove Expansion). Specifically, climate and governance indicators are 11 and 5 more in Cluster 1 than in Cluster 2, while socio-economic indicators are nine fewer in Cluster 1 ( Fig. 2 ). This suggests that climate change has a higher synergy with mangrove loss, while socio-economic development shows a stronger synergy with mangrove expansion. Fig. 2. Open in a new tab Distribution of climate, socio-economic, and governance indicators in mangrove expand and loss clusters. The bar chart illustrates the number of climate, socio-economic, and governance indicators in mangrove loss (Cluster 1) and mangrove expansion (Cluster 2). Red bars represent Cluster 1, while blue bars represent Cluster 2. The circular graph displays the proportional distribution of these indicators across the two clusters, where C, S, and G denote climate, socio-economic, and governance indicators, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) On one hand, the frequent occurrence of extreme climate events exacerbates mangrove loss. Indicators M-1 (mangrove loss caused by erosion) to M-9 (mangrove gross loss) form the core of Cluster 1, exhibiting strong proximity (≥ 0.8) with C-10 (number of climate-related disasters) and C-12 (tropical storm frequency). There is also a strong correlation (≥ 0.8) between C-10, C-12, and C-4 (CO 2 emission targets excluding land use) and C-9 (annual net greenhouse gas emission targets). Additionally, according to the Space results, mangrove loss has a high proximity (≥ 0.75) with C-4 to C-9. The accumulation of greenhouse gases in the atmosphere drives global warming, which in turn intensifies extreme events such as tropical storms, floods, and droughts—further degrading mangrove habitats and accelerating their decline. Therefore, controlling and reducing greenhouse gas emissions is critical to mitigating extreme climate risks and improving mangrove ecosystem resilience. On the other hand, mangrove degradation contributes to rising atmospheric greenhouse gas concentrations. Mangroves possess exceptional carbon sequestration capacity, absorbing CO₂ through biological growth and microbial activity, and storing it in coastal sediments [ 48 , 49 ]. As one of the largest blue carbon reservoirs globally, mangroves account for > 50% of the total carbon stock in the world’s blue carbon ecosystems [ 50 ]. Their loss not only diminishes the absorption of atmospheric CO₂, but also releases long-stored carbon back into the atmosphere through processes such as burning and land-use conversion. Therefore, the intensification of mangrove loss will not only weaken the carbon sequestration capacity of natural ecosystems but may also cause mangroves to shift from carbon sinks to carbon sources, accelerating the rise in atmospheric greenhouse gas concentrations and triggering a vicious cycle of global climate change. 3.2. Mangrove restoration efforts promote national economic development and reflect government efficiency In Cluster 2, M-13 (mangrove area) exhibits strong connections with socio-economic indicators such as S-9 (gross capital formation), S-24 (installed capacity of non-renewable energy), and S-26 (total trade of low-carbon products). Through these relationships, M-13 is further linked to broader economic and demographic indicators, including S-2 (urban population), S-4 (total population), S-7 (GDP), and S-13 (agricultural output). These indicators are positioned at the core of the cluster and serve as hubs that connect with other socio-economic variables. Enhancing their performance can generate positive cascading effects across the socio-economic system. This suggests that increasing mangrove protection can contribute to both economic development and social stability. Additionally, M-13 (mangrove area) is closely linked to G-2 (government efficiency) and G-7 (economic complexity index) in Cluster 3, indicating that a country’s mangrove management and restoration performance can serve as a proxy for both its government efficiency and economic development. Furthermore, M-13 (mangrove area), G-2 (government efficiency), and G-7(economic complexity index) exhibit high betweenness centrality and occupy central positions in the Space network. Enhancing these indicators can further empower the development of other network indicators. On one hand, improving legal frameworks to protect G-5 (area of Ramsar wetlands) and enforcing strict penalties for illegal activities such as deforestation and fishing, can effectively enhance M-10 (mangrove gross gain). On the other hand, accelerating the adoption of S-22 (installed capacity of renewable energy generation) and reducing the use of S-23 (electricity generation of nonrenewable energy) to lower greenhouse gas emissions can help mitigate the worsening of mangrove loss. 3.3. Improving “bridge” indicators is crucial for enhancing mangrove condition and addressing climate change In the Space network, the indicators with the highest betweenness centrality are G-6 (number of Ramsar wetland), M-13 (mangrove area), G-2 (government efficiency), C-3 (Standardized Precipitation Evapotranspiration Index), C-11 (temperature change), and C-1 (sea level rise), with G-6 in Cluster 3 ranking significantly higher than all other indicators (Fig. S3b, Table S1). Establishing wetland protected areas is essential for reducing mangrove degradation and promoting ecological restoration, while efficient government governance is the key to achieving this goal. First, wetland protected areas can minimize external disturbances to mangrove ecosystems. Accelerating the natural recovery by replanting and restoring mangroves (M-10) in degraded areas leads to increased mangrove area (M-13) and forest cover (M-15), while enhancing the carbon sequestration capacity of the ecosystem (M-14). Second, wetland protection can effectively prevent destructive activities such as deforestation, pollution, and land reclamation. It helps mitigate mangrove loss caused by human activities such as international trade (M-4), while enhancing ecosystem resilience to extreme climate disasters (C-10), thereby reducing mangrove loss from natural erosion (M-1) and extreme disasters (M-2). At the same time, improving government efficiency (G-2) can influence multiple nodes, producing an amplification effect. G-6 (number of Ramsar wetlands) functions as a “bridge” between Cluster 1 and Cluster 2, connecting climate and socio-economic indicators, thereby contributing to the formation of a stable network framework. Specifically, G-6 is strongly connected to G-2 (government efficiency), G-5 (area of Ramsar wetlands), G-7 (economic complexity index), C-3 (Standardized Precipitation Evapotranspiration Index), and S-11 (percentage of industrial added value in GDP). Notably, both G-2 and C-3 are high-betweenness centrality indicators and exhibit synergistic relationships with C-1 and C-11. These connections facilitate the integration of climate and socio-economic indicators within the network, such as C-2 (river sediment capture index), S-10 (percentage of gross capital formation in GDP), and S-19 (net foreign investment outflow as a share of GDP). Enhancing governance capacity and economic complexity can therefore help reduce mangrove loss, while improving socio-economic and climate outcomes across both Cluster 1 and Cluster 2. Additionally, mangrove restoration exerts a reciprocal effect on both climate and socio-economic systems. First, improving the ecosystem’s carbon sequestration capacity (M-14) helps achieve greenhouse gas net emission targets (C-9), thereby controlling changes in global average temperature (C-11) and precipitation index (C-3), reducing the rate of sea level rise (C-1), and mitigating the occurrence of tropical storms (C-12) and extreme climate disasters (C-10). Furthermore, the stability of the climate system not only creates conditions for industrialization (S-11) but also improves future expectations, thereby boosting foreign investment (S-17) and gross capital formation (S-9), which in turn drives increases in the electricity installed capacity of renewable energy (S-22) and GDP (S-7) (Fig. S3a). 3.4. Complex and influential indicators are separated in mangrove expansion and loss clusters To objectively identify the key indicators of mangrove systemic risk, we incorporate Complexity and Eigenvector Centrality measurements, together with betweenness centrality, to construct a multidimensional evaluation system, thus assisting policymakers with specific policy needs. Based on the proposed evaluation system and network analysis, indicators with high eigenvector centrality should be prioritized in climate policy decision-making. Key indicators—such as mangrove gross loss (M-9), mangrove loss driven by international trade (M-4), and net CO₂ emission targets (C-8) exhibit high eigenvector centrality ( Fig. 3 a). Enhancing these indicators can generate positive cascading effects, facilitating the coordinated development across related domains. This is particularly relevant for environmental indicators in Mangrove Loss (Cluster 1), which directly influence critical areas such as biodiversity and habitat (G-4), carbon sequestration capacity (M-14), and extreme climate disasters (C-10). Prioritizing these indicators can substantially increase policy efficiency and support broader improvements in climate and economic systems. Meanwhile, complex indicators such as the installed capacity of renewable energy generation (S-22), GDP (S-7), and artificial surfaces (S-1) in Mangrove Expand (Cluster 2) reflect latent systemic vulnerabilities. Advancing these areas requires multi-dimensional support, which means the systemic risk can result in irreversible trauma. Fig. 3. Open in a new tab (a) Eigenvector centrality and complexity measurements of indicators in space network. The color of the nodes represents the complexity of the indicators—darker colors indicate higher complexity—while the size of the nodes reflects eigenvector centrality, with larger nodes denoting greater systemic influence. The color of the background partition represents different clusters and is consistent with ( Fig. 1 ). Indicators with higher eigenvector centrality are primarily concentrated in the Mangrove Loss (Cluster 1), whereas indicators with higher complexity tend to appear in the Mangrove Expand (Cluster 2). The rankings of these two types of indicators do not overlap (Fig. S1) . A Kendall’s Tau correlation between indicator complexity and eigenvector centrality yielded a value of −0.66 ( p < 0.001), indicating a strong negative association ( Table S2 ). (b) Eigenvector Centrality of various countries. The color of the regions represents the eigenvector centrality of the countries—darker colors indicate higher eigenvector centrality. (c) Complexity of multiple countries. The color of the regions represents the complexity of the countries—darker colors indicate higher complexity. It is important to note that mangrove ecosystems occupy the intersection of two key indicator types: they are both the target of high eigenvector centrality indicators and function as ecological protective buffers for high-complexity indicators. This dual role highlights the strategic importance of mangroves in advancing climate change mitigation and socio-economic development. Their degradation could initiate cascading crises, including the collapse of carbon sinks and the loss of biodiversity. Scientific conservation efforts for mangroves can reinforce a positive feedback mechanism between renewable energy transitions and sustainable economic growth. The analysis indicates that mangrove loss is closely linked to climate change and socio-economic development, forming a bidirectional vicious cycle. Rising greenhouse gas emissions exacerbate mangrove loss by intensifying extreme weather events such as tropical storms. In turn, mangrove loss reduces the ecosystem’s carbon sequestration capacity, converting carbon sinks into sources and accelerating global warming. Moreover, mangrove conservation is strongly associated with GDP, economic transformation, investment, and government efficiency. Strengthening mangrove conservation not only promotes economic growth but also reflects the quality of governance. Therefore, enhancing government efficiency and reinforcing wetland protection are critical for effective mangrove conservation. Legislative measures should be implemented to minimize human interference and improve the resilience of mangrove ecosystems. Policymakers should incorporate mangrove protection into national climate and development strategies, using it as a tool to strengthen environmental indicators with high eigenvector centrality while establishing an ecological safety buffer for the stable development of high-complexity socio-economic systems, thereby achieving the dual strategic goals of enhancing climate resilience and promoting green economic growth. 4. Discussion This study contributes a novel approach to understanding the systemic role of mangrove ecosystems. By integrating ecological, socio-economic, and policy indicators, it demonstrates how mangrove loss or restoration can either compound risks or catalyze synergistic development. The proposed Mangrove Multisystemic Risk Space network reveals the complementary relationships among mangroves, climate change, socio-economic development, and governance performance. Using betweenness centrality, the study identifies key indicators that exert cross-dimensional influence within the network. Through eigenvector centrality and complexity analysis, it highlights core indicators that drive systemic dynamics or reflect vulnerability to systemic risk. Building on this, the study further evaluates eigenvector centrality and complexity at the regional level. By combining complex network analysis with a multidimensional evaluation framework, this research provides a robust scientific foundation for informing policies that balance mangrove conservation, climate change mitigation, and socio-economic development. At the regional level, there is a moderate positive correlation between eigenvector centrality and complexity ( Fig. 4 b). Countries with high eigenvector centrality typically possess well-developed core indicators that can catalyze the advancement of others, thereby influencing the entire system. However, certain countries show imbalances between complexity and centrality. Fig. 4. Open in a new tab Mangrove multisystemic risk performance space in selected countries and regional eigenvector-complexity relationship. (a) The Mangrove Multisystemic Risk Performance Space includes eight countries: Panama (PAN), New Zealand (NZL), India (IND), the United States (USA), Indonesia (IDN), Brazil (BRA), Australia (AUS), and Malaysia (MYS). Among these, PAN and NZL exhibit the highest eigenvector centrality, while the USA and IND show strong performance in complexity. IDN, BRA, AUS, and MYS are among the world’s largest mangrove-holding countries. Node colors represent the sustainability score of each indicator. (b) Relationship between regional eigenvector centrality (x-axis) and complexity (y-axis). Node size corresponds to each country’s mangrove area, and node color reflects income level based on World Bank classifications. Countries with eigenvector centrality > 0.065 and complexity > 50 are categorized as “influential” and “complex” countries, respectively. The Kendall’s Tau coefficient between eigenvector centrality and complexity is 0.51 ( p < 0.001), indicating a significant positive correlation (Table S2). 4.1. Dual policy priorities on complexity enhancement and “bridge” indicators promotion in Panama and New Zealand Panama (PAN) is a small open economy with limited natural resources, whose income mainly relies on its strategic location as a global trade hub in international trade [ 51 ]. On the contrary, New Zealand (NZL) is a developed country with comprehensive social welfare, where rich natural resources can boost the agriculture and tourism economy [ 52 ]. PAN and NZL both exhibit high eigenvector centrality but low complexity, suggesting that while they perform well in different influential indicators, they lag in more complex indicators. Consequently, their capacity to withstand systemic risk may require further strengthening. Two strategic alternatives can be considered for them. On the one hand, both countries can focus on enhancing complex indicators to strengthen industrial performance. For example, by improving indicators such as S-1 (artificial surface), S-9 (gross capital formation), and S-23 (electricity generation from nonrenewable energy), and ensuring adequate labor supply via S-4 (population) and S-2 (urban population), economic growth can be stimulated. Concurrently, enhancing ecological conditions by improving G-5 (area of Ramsar wetlands) and M-15 (forest area) can further support tourism and environmental resilience. Progress in these foundational areas lays the groundwork for advancing higher-complexity indicators such as S-7 (GDP), S-21 (electricity generation from renewable energy), and S-26 (total trade in low-carbon technology products), thereby gradually raising the overall complexity level of the region. On the other hand, emphasis should be placed on strengthening “bridge” indicators with low sustainability scores. Targeted improvements in these areas can generate spillover effects, facilitating more balanced and coordinated development. For instance, both New Zealand and Panama score poorly on G-6 (number of Ramsar wetlands). Establishing wetland conservation zones and expanding mangrove coverage can not only reduce anthropogenic disturbances to mangrove ecosystems but also enhance the carbon sequestration capacity through “bridge” indicators like M-14 (ecosystem carbon storage). Moreover, Panama’s rising coastlines and reliance on global trade necessitate a governance-focused approach, where strengthening G-2 (government effectiveness) is as critical as improving C-1 (sea level rise). Through protecting industrial infrastructure and human settlements, while enhancing administrative efficiencies, Panama can ensure the effective implementation of ecological protection and economic development policies. 4.2. Strategic reinforcement of high-centrality emissions control and governance “bridge” indicators in global economic hubs India (IND), the United States (USA), and Brazil (BRA), as major global economies, generally possess more advanced industrial systems than other countries. They demonstrate strong performance in high-complexity indicators such as S-21 (electricity generation from renewable sources) and S-22 (installed capacity of renewable energy). These countries also outperform others in fundamental economic indicators, including S-1 (artificial surface), S-7 (GDP), and S-9 (gross capital formation). Although they share above similarities in resource capacity and production ability, their governance and socio-economic development vary significantly. India’s multi-level governance system often leads to uneven policy implementation across regions. Brazil’s reliance on commodity exports is limited with corruption and political instability. The United States has advantage in technological and financial capacity, but its federal–state system is characterized in fragmented and delayed climate policy responses, which undermines the progress of SDGs. These institutional and structural differences highlight the need for country-specific governance priorities to mitigate systemic risk. These differences underscore the importance of tailoring governance priorities. On the one hand, the focus is on indicators with high eigenvector centrality. For instance, the greenhouse gas emissions of the United States and India—represented by C-4 (CO2 emission targets excluding land use) and C-5 (CO2 emission targets including land use) remain poor, with their combined CO₂ emissions accounting for approximately 23% of the global total [ 53 ]. Rising global greenhouse gas concentrations have already led to extreme heat and drought, which are hampering agricultural productivity and intensifying India’s food insecurity [ 54 ]. The United States is also among the countries most affected by climate change. In 2023 alone, heatwaves, wildfires, storms, and floods caused economic losses exceeding USD 1 billion [ 55 ]. Reducing greenhouse gas emissions would not only directly mitigate agricultural and economic losses from extreme climate disasters (C-10) and tropical storms (C-12), but also generate cascading benefits across related indicators such as C-8 (net CO₂ emission target), M-9 (mangrove gross loss), S-17 (foreign investment), and G-4 (biodiversity and habitat), thereby strengthening national resilience against systemic risk. On the other hand, utilizing “bridge” indicators located at the periphery of clusters is equally critical for preventing systemic risk. In the case of the United States, future policy efforts should prioritize controlling temperature rise (C-11) and precipitation variability (C-3), alongside expanding mangrove area (M-13) to enhance the climate regulation capacity. For India, the development disparity between urban and rural areas, as well as the industrial structure, requires an enhancement in G-7 (economic complexity index). Because of the reliance on the export of primary products and a weak industrial foundation, Brazil needs to improve S-11 (percentage of industrial added value in GDP). Moreover, strengthening G-2 (government effectiveness) alongside efforts to reduce corruption (G-1) is essential for both India and Brazil. By improving these low-performing yet influential “bridge” indicators, it is possible to activate cascading benefits that support progress in other dimensions, such as C-1 (sea level rise), S-10 (percentage of gross capital formation in GDP), and S-26 (total trade in low-carbon technology products). Ultimately, this approach can enable the synergistic optimization of ecological and economic objectives, fostering a more resilient trajectory toward sustainable development. 4.3. Synergizing mangrove conservation policies with low-carbon trade metrics in Australia’s climate resilience strategy As the country with the third-largest mangrove area in the world, Australia has made notable progress in recent years by increasing fiscal investment to expand wetland coverage (G-6) and partially reduce mangrove loss caused by human activities (M-3). However, total mangrove loss (M-9) and net loss (M-8) remain at relatively high levels ( Fig. 4 a), and the effectiveness of current policies in protecting marine biodiversity has been limited [ 56 ]. Given the global impact of climate change, the Australian government must adopt a more proactive stance in international climate initiatives to address the potential systemic environmental risks posed by mangrove degradation. According to the complex network analysis, priority should be given to key indicators with high eigenvector centrality but low sustainability scores. For example, enhancing G-9 (marine staff capacity) could reduce the anthropogenic mangrove loss and drive broader marine ecosystem conservation. In parallel, strengthening S-14 (percentage of agriculture added value in GDP), S-17 (foreign investment), and mitigating tropical storms (C-12) would contribute to improving G-7 (economic complexity index). This would support a high-quality economic development, while simultaneously reducing mangrove degradation linked to industrial pollution and extreme weather events. When considering Australia’s export-oriented economy and vulnerability to global market fluctuations, strengthening Cluster 2 indicators with high betweenness centrality holds significant potential to improve the overall effectiveness of mangrove conservation and restoration. Notably, S-26 (total trade of low-carbon products) and M-10 (mangrove gross gain) exhibit strong bridging and diffusion potential. Improving these indicators can generate synergistic progress in related domains, such as C-5 (CO 2 emission targets including land use), which have high eigenvector centrality, and also promote the development of high-complexity indicators like S-9 (gross capital formation) and S-21 (electricity generation from renewable energy). Compared to fossil fuels, the use of renewable energy sources—such as solar, wind, and hydropower—not only significantly reduces greenhouse gas emissions but also avoids the release of harmful pollutants like sulfur dioxide (SO 2 ), nitrogen oxides (NO X ), and fine particulate matter (PM2.5), which further benefits public health S-2 (urban population). This approach aligns with Australia’s ambition in environmental management while addressing the systemic risks associated with mangrove degradation. 5. Conclusion This study constructed the Mangrove Multisystemic Risk Space to uncover the transmission mechanisms of systemic risk embedded within mangrove ecosystems. It conducted an in-depth analysis of the interactions among ecological, climatic, and socio-economic systems, and identified high-leverage indicators that exert cross-dimensional influence. Additionally, a multidimensional evaluation framework was established to distinguish between indicators that initiate cascading effects and those that remain dependent on external conditions. By integrating complex network analysis within this multidimensional framework, the study seeks to inform tailored, evidence-based policy interventions aimed at enhancing systemic risk resilience in mangrove governance. The complex network comprising 64 indicators spanning four dimensions—mangrove ecosystems, climate change, socio-economic development, and governance, to elucidate the intricate interconnections among them. By integrating the complex network with the proposed multidimensional evaluation framework, we highlighted the pivotal role of the mangrove ecosystem in mitigating systemic risk and enabling synergistic promotion across climate and socio-economic domains. Mangrove degradation triggers cascading risks, eroding carbon sequestration capacity and undermining biodiversity while destabilizing socio-economic systems. Conversely, mangrove expansion correlates with socio-economic resilience and governance efficiency, serving as both a climate buffer and a proxy for systemic sustainability. Leveraging the results from the complex network and multidimensional evaluation framework, we derived tailored policy priorities for countries based on their unique systemic profiles. Nations exhibiting high influence but low complexity—such as Panama and New Zealand—would benefit from promoting complex indicators such as renewable energy usage and institutional integration effectiveness in mangrove restoration efforts. In contrast, countries with high complexity, such as the United States, Brazil, and India, can focus on strengthening their influential indicators, including mitigating extreme climate events and reducing greenhouse gas emissions. Specifically, policymakers can prioritize improvements in indicators either with high eigenvector and betweenness centrality—such as Ramsar Site coverage, ecosystem protection and administrative governance since they can produce ripple effects across multidimensional sustainability, or with high complex metrics, such as renewable energy usage, green economic development, and technology shift since they can enhance the vulnerability in facing systemic environmental risk. Strategic investment in these leverage points can accelerate progress in climate action, socio-economic development, and systemic risk mitigation concurrently. In Australia’s case, expanding marine management training and natural ecosystem protection are critical to counteract the impacts of industrial pollution and extreme weather events. policy incentives should also accelerate the adoption of low-carbon technologies and renewable energy to reduce greenhouse gas emissions and enable a high-quality socio-economic transition. A limitation of this study is that, although the Mangrove Multisystemic Risk Space facilitates coordinated development among ecological, environmental, and socio-economic systems and reveals synergistic relationships, it does not account for the potential trade-offs that may arise from advancing these indicators. For instance, while the widespread promotion of renewable energy technologies and low-carbon equipment can help alleviate the escalating climate crisis, such initiatives may also lead to industrial disruptions, such as land-use conflicts caused by the construction of photovoltaic facilities, or market fluctuations resulting from increased subsidies for low-carbon products. Therefore, future research should incorporate policy implementation challenges driven by market dynamics to more comprehensively address systemic risk governance. Data and code availability Climate indicators were obtained from the International Monetary Fund’s Climate Change Indicators Dashboard ( https://climatedata.imf.org/ ). Socioeconomic data were sourced from the World Bank’s World Development Indicators database ( https://databank.worldbank.org/source/world-development-indicators#advancedDownloadOptions ). Mangrove loss footprint data were derived from our previous study ( https://www.scilit.com/publications/ecb09464d6cbb849a1545f0eaa8cce53 ) [ 57 ]. To capture a comprehensive understanding of the mangrove landscape, encompassing drivers and hotspots of both loss and gain, we incorporated a wide range of metrics consolidated from various sources, as detailed in the reference study ( https://www.nature.com/articles/s41467–022–33962-x#data-availability ) [ 22 ]. All code and raw data used in this study are publicly available at the following GitHub repository: https://github.com/lwt852/MangroveRoleinMitigatingSystemicRisk.git . Interested readers are encouraged to download the materials directly from this source. CRediT authorship contribution statement Mimi Gong: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Ke Yu: Writing – review & editing, Visualization, Methodology. Qiang Huang: Writing – review & editing, Visualization, Validation, Methodology, Formal analysis, Data curation. Yinglan A: Writing – review & editing, Visualization. Miriam Aczel: Writing – review & editing, Visualization. Ye Li: Writing – review & editing, Methodology. Maofang Ran: Writing – review & editing, Visualization, Methodology. Yan Cheng: Visualization. Kaiji Li: Methodology. Shen Qu: Writing – review & editing, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Declaration of competing interest The authors declare that they have no conflicts of interest in this work. Acknowledgments Shen Qu thanks the financial support from the National Natural Science Foundation of China (52425005). Mimi Gong thanks the support from William W. and Evelyn M. Taylor Endowed Fellowship. Biographies Mimi Gong , Assistant Professor at Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University, specializes in mangrove conservation, coupled forest–human–land systems, Sustainable Development Goals, and wetland basin management. She has published over ten SCI-indexed papers—in multidisciplinary journals spanning geographic information, ecology , and environmental management —ten of which are in the JCR Q1 zone. Her research has yielded practical outcomes, including mapping the global footprint of mangrove loss and analyzing spatial drivers, as well as constructing a complex-network feedback framework, applying machine learning to assess its spatiotemporal stability and derive policy insights. She is the recipient of the Michigan State University Doctoral Dissertation Defense Fellowship, the Serf Lawrence Conservation Ecology Scholarship, the Williams Taylor Sustainable Development Award, and China’s National Scholarship. Her future endeavors will focus on exploring national-scale mangrove conservation strategies through the lenses of national responsibility and equity from the standpoint of the global mangrove loss footprint, and using complex network modeling to investigate potential development strategies for sustainable development, mangrove protection, and governance of the Yellow River Basin, providing policy recommendations for decision-makers at different administrative scales. Qiang Huang is a PhD candidate in the Center for Energy & Environmental Policy Research at Beijing Institute of Technology. He received a bachelor degree of economics from the Sichuan Agricultural University in China, in 2023. His research interests include Mangrove Protection, Environmental Systemic Risk, and Artificial Intelligence Modeling. Shen Qu ( BRID: 07758.00.65053 ) is a professor in the Center for Energy & Environmental Policy Research, Beijing Institute of Technology. His “Environmental Data Science and Technology” project was funded by the National Science Fund for Distinguished Young Scholars. His research interests include environmental systems engineering, environmental management, and related data science methods. He has published over 90 papers in prestigious journals such as Nature Communications, Global Environmental Change, Environmental Science & Technology, Engineering , and Fundamental Research . His achievements have been applied in various aspects, including China’s first environmental pollution liability insurance risk control system, economic and ecological impact assessment of the national water grid project, and demonstration projects for carbon neutrality in urban sewage systems. He serves as the deputy secretary-general and executive director of the Chinese Society of Energy Economics and Management, as well as the deputy director of the Youth Scientists Branch of the Chinese Society for Environmental Science. He is the associate editors of Resources, Conservation & Recycling , the deputy editor of the Journal of Cleaner Production , and an editorial board member of the Journal of Beijing Institute of Technology (Social Sciences Edition) . Footnotes Peer review under the responsibility of Editorial Board of Fundamental Research. Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.fmre.2025.11.003 . Contributor Information Qiang Huang, Email: [email protected]. Shen Qu, Email: [email protected]. Appendix. Supplementary materials mmc1.docx (328.6KB, docx) References 1. Cheng D. Socioeconomic and political interactions. Nat. Clim. Chang. 2024;14:1223. [ Google Scholar ] 2. OECDOECD economic outlook, 2025. Interim Report March 2025 https://www.oecd.org/en/publications/oecd-economic-outlook-interim-report-march-2025_89af4857-en.html (accessed 24 September 2025). 3. Wang K., Liu J., Xia J., et al. Understanding the impacts of climate change and socio-economic development through food-energy-water nexus: A case study of Mekong River Delta. Resour. Conserv. Recycl. 2021;167 [ Google Scholar ] 4. Battiston S., Mandel A., Monasterolo I., et al. A climate stress test of the financial system. Nat. Clim. Chang. 2017;7:283–288. [ Google Scholar ] 5. IPCCClimate change 2023: synthesis report. Contribution of working groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on climate change, 2023. https://www.ipcc.ch/report/sixth-assessment-report-cycle/ (accessed 23 March 2025). 6. Richards D.R., Thompson B.S., Wijedasa L. Quantifying net loss of global mangrove carbon stocks from 20 years of land cover change. Nat. Commun. 2020;11:4260. doi: 10.1038/s41467-020-18118-z. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Bi J., Yang J., Liu M., et al. Toward systemic thinking in managing environmental risks. Engineering. 2021;7:1518–1522. [ Google Scholar ] 8. Yang J. Dealing with systemic environmental risks. Nat. Sustain. 2025;8:466. [ Google Scholar ] 9. Kareem A., Parveen K.H., Sharma S.V.K., et al. Mangroves in environmental engineering: Harnessing the multifunctional potential of nature’s coastal architects for sustainable ecosystem management. Results Eng. 2024;21 [ Google Scholar ] 10. Montgomery J.M., Bryan K.R., Mullarney J.C., et al. Attenuation of storm surges by coastal mangroves. Geophys. Res. Lett. 2019;46:2680–2689. [ Google Scholar ] 11. Wang Y.-S., Gu J.-D. Ecological responses, adaptation and mechanisms of mangrove wetland ecosystem to global climate change and anthropogenic activities. Int. Biodeterior. Biodegrad. 2021;162 [ Google Scholar ] 12. GMAThe state of the world’s mangroves, 2022. https://www.wetlands.org/publication/the-state-of-the-worlds-mangroves-2022/ (accessed 3 April 2025). 13. IUCNRed list of ecosystems, 2024. https://iucn.org/resources/conservation-tool/iucn-red-list-ecosystems (accessed 3 April 2025). 14. Atwood T.B., Connolly R.M., Almahasheer H., et al. Global patterns in mangrove soil carbon stocks and losses. Nat. Clim. Chang. 2017;7:523–528. [ Google Scholar ] 15. Morrissette H.K., Baez S.K., Beers L., et al. Belize Blue carbon: Establishing a national carbon stock estimate for mangrove ecosystems. Sci. Total Environ. 2023;870 doi: 10.1016/j.scitotenv.2023.161829. [ DOI ] [ PubMed ] [ Google Scholar ] 16. GMAThe state of the world’s mangroves, 2024. https://www.mangrovealliance.org/mangrove-forests/ (accessed 27 May 2025) . 17. Zu Ermgassen P.S.E., Mukherjee N., Worthington T.A., et al. Fishers who rely on mangroves: Modelling and mapping the global intensity of mangrove associated fisheries. Estuar. Coast. Shelf Sci. 2020;247 [ Google Scholar ] 18. Bastien-Olvera B.A., Rivera A., Gray E., et al. Mangrove preservation could have significantly reduced damages from Hurricane Otis on the coast of Guerrero, Mexico. Sci. Total Environ. 2024;957 doi: 10.1016/j.scitotenv.2024.177822. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Menéndez P., Losada I.J., Torres-Ortega S., et al. The global flood protection benefits of mangroves. Sci. Rep. 2020;10:4404. doi: 10.1038/s41598-020-61136-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. van Zelst V.T.M., Dijkstra J.T., van Wesenbeeck B.K., et al. Cutting the costs of coastal protection by integrating vegetation in flood defences. Nat. Commun. 2021;12:6533. doi: 10.1038/s41467-021-26887-4. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Goldberg L., Lagomasino D., Thomas N., et al. Global declines in human-driven mangrove loss. Glob. Change Biol. 2020;26:5844–5855. doi: 10.1111/gcb.15275. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Hagger V., Worthington T.A., Lovelock C.E., et al. Drivers of global mangrove loss and gain in social ecological systems. Nat. Commun. 2022;13:6373. doi: 10.1038/s41467-022-33962-x. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Saintilan N., Khan N.S., Ashe E., et al. Thresholds of mangrove survival under rapid sea level rise. Science. 2020;368:1118–1121. doi: 10.1126/science.aba2656. [ DOI ] [ PubMed ] [ Google Scholar ] 24. Sippo J.Z., Lovelock C.E., Santos I.R., et al. Mangrove mortality in a changing climate: An overview. Estuar. Coast. Shelf Sci. 2018;215:241–249. [ Google Scholar ] 25. Belaïd F., Unger C. Crafting effective climate, energy, and environmental policy: Time for action. Humanit. Soc. Sci. Commun. 2024;11:5. [ Google Scholar ] 26. Kammerlander A., Schulze G.G. Political economic correlates of environmental policy. Environ. Res. Lett. 2021;16 [ Google Scholar ] 27. Wang Z., Zhang Y., Kang S., et al. Research progresses of microplastic pollution in freshwater systems. Sci. Total Environ. 2021;795 doi: 10.1016/j.scitotenv.2021.148888. [ DOI ] [ PubMed ] [ Google Scholar ] 28. Orchard S.E., Stringer L.C., Quinn C.H. Mangrove system dynamics in Southeast Asia: Linking livelihoods and ecosystem services in Vietnam. Reg. Environ. Change. 2016;16:865–879. [ Google Scholar ] 29. Yang X., Meng F., Fu P., et al. Spatiotemporal change and driving factors of the eco-environment quality in the Yangtze River Basin from 2001 to 2019. Ecol. Indic. 2021;131 [ Google Scholar ] 30. Coletta V.R., Pagano A., Zimmermann N., et al. Socio-hydrological modelling using participatory system dynamics modelling for enhancing urban flood resilience through blue-green infrastructure. J. Hydrol. 2024;636 doi: 10.1016/j.jhydrol.2024.131248. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Turschwell M.P., Tulloch V.J.D., Siegers M., et al. Multi-scale estimation of the effects of pressures and drivers on mangrove forest loss globally. Biol. Conserv. 2020;247 [ Google Scholar ] 32. Lawrence J., Blackett P., Cradock Henry N.A. Cascading climate change impacts and implications. Clim. Risk Manag. 2020;29 doi: 10.1016/j.mex.2020.100893. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Xue R., Wang C., Zhao L., et al. Agricultural intensification weakens the soil health index and stability of microbial networks. Agric. Ecosyst. Environ. 2022;339 [ Google Scholar ] 34. Hidalgo C.A., Klinger B., Barabási A.-L., et al. The product space conditions the development of nations. Science. 2007;317:482–487. doi: 10.1126/science.1144581. [ DOI ] [ PubMed ] [ Google Scholar ] 35. Rigby D.L. Technological relatedness and knowledge space: Entry and exit of US cities from patent classes. Reg. Stud. 2015;49(11):1922–1937. [ Google Scholar ] 36. Ma F., Wang H., Schandl H., et al. Exploring the relationship between economic complexity and resource efficiency. Resour. Conserv. Recycl. 2022;186 [ Google Scholar ] 37. Qi X. Building a bridge between economic complexity and the blue economy. Ocean Coast. Manag. 2022;216 [ Google Scholar ] 38. Ma F., Wang H., Tzachor A., et al. The disparities and development trajectories of nations in achieving the sustainable development goals. Nat. Commun. 2025;16:1107. doi: 10.1038/s41467-025-56076-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Mealy P., Teytelboym A. Economic complexity and the green economy. Res. Policy. 2022;51 [ Google Scholar ] 40. Gong M., Teller N., Golebie E.J., et al. Unveiling complementarities between mangrove restoration and global sustainable development goals. J. Clean. Prod. 2024;474 [ Google Scholar ] 41. Gong M., Yu K., Zhou C., et al. SDG space: Revealing the structure and complementarities among sustainable development goals in China. Fundam. Res. 2025;5:360–369. doi: 10.1016/j.fmre.2024.01.005. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Gong M., Yu K., Xu Z., et al. Unveiling complementarities between national sustainable development strategies through network analysis. J. Environ. Manage. 2024;350 doi: 10.1016/j.jenvman.2023.119531. [ DOI ] [ PubMed ] [ Google Scholar ] 43. Goldberg L., Lagomasino D., Thomas N., et al. Global mangrove loss extent, land cover change, and loss drivers, 2000-2016. ORNL Distr. Act. Arch. Center. 2022;v1 doi: 10.3334/ORNLDAAC/1768. [ DOI ] [ Google Scholar ] 44. Newman M.E.J. Modularity and community structure in networks. Proc. Natl. Acad. Sci. USA. 2006;103:8577–8582. doi: 10.1073/pnas.0601602103. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Blondel V.D., Guillaume J.-L., Lambiotte R., et al. Fast unfolding of communities in large networks. J. Stat. Mech. 2008 [ Google Scholar ] 46. Brandes U. A faster algorithm for betweenness centrality. J. Math. Sociol. 2001;25:163–177. [ Google Scholar ] 47. Bastian M., Heymann S., Jacomy M. Gephi: An open source software for exploring and manipulating networks. Proc. Int. AAAI Conf. Weblogs Soc. Media. 2009;3:361–362. [ Google Scholar ] 48. Friess D.A., Rogers K., Lovelock C.E., et al. The state of the world’s mangrove forests: Past, present, and future. Annu. Rev. Environ. Resour. 2019;44:89–115. [ Google Scholar ] 49. Zhu J.-J., Yan B. Blue carbon sink function and carbon neutrality potential of mangroves. Sci. Total Environ. 2022;822 doi: 10.1016/j.scitotenv.2022.153438. [ DOI ] [ PubMed ] [ Google Scholar ] 50. Siikamäki J., Sanchirico J.N., Jardine S.L. Global economic potential for reducing carbon dioxide emissions from mangrove loss. Proc. Natl. Acad. Sci. USA. 2012;109:14369–14374. doi: 10.1073/pnas.1200519109. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. IISSGreat-power competition in the Panama Canal, 2025. https://www.iiss.org/online-analysis/online-analysis/2025/04/great-power-competition-in-the-panama-canal/ (accessed 16 May 2025). 52. Paul T., Kimberley M.O., Beets P.N. Natural forests in New Zealand—a large terrestrial carbon pool in a national state of equilibrium. For. Ecosyst. 2021;8:34. [ Google Scholar ] 53. IMFClimate change indicators dashboard, 2024. https://climatedata.imf.org/ (accessed 24 September 2025). 54. Chuphal D.S., Kushwaha A.P., Aadhar S., et al. Drought atlas of India, 1901–2020. Sci. Data. 2024;11:7. doi: 10.1038/s41597-023-02856-y. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 55. NOAAU.S. struck with historic number of billion-dollar disasters in 2023, 2024. https://www.noaa.gov/news/us-struck-with-historic-number-of-billion-dollar-disasters-in-2023 (accessed 24 September 2025). 56. Davey M., Watson J.E.M. Australia’s inadequate marine protection. Science. 2025;387:586–587. doi: 10.1126/science.adt7848. [ DOI ] [ PubMed ] [ Google Scholar ] 57. Qu S., Gong M., Wang G., et al. Global mangrove loss footprint mappings across space and time. Res. Square. 2025 doi: 10.21203/rs.3.rs-5778596/v1. Preprint at. [ DOI ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials mmc1.docx (328.6KB, docx) Data Availability Statement Climate indicators were obtained from the International Monetary Fund’s Climate Change Indicators Dashboard ( https://climatedata.imf.org/ ). Socioeconomic data were sourced from the World Bank’s World Development Indicators database ( https://databank.worldbank.org/source/world-development-indicators#advancedDownloadOptions ). Mangrove loss footprint data were derived from our previous study ( https://www.scilit.com/publications/ecb09464d6cbb849a1545f0eaa8cce53 ) [ 57 ]. To capture a comprehensive understanding of the mangrove landscape, encompassing drivers and hotspots of both loss and gain, we incorporated a wide range of metrics consolidated from various sources, as detailed in the reference study ( https://www.nature.com/articles/s41467–022–33962-x#data-availability ) [ 22 ]. All code and raw data used in this study are publicly available at the following GitHub repository: https://github.com/lwt852/MangroveRoleinMitigatingSystemicRisk.git . Interested readers are encouraged to download the materials directly from this source. Articles from Fundamental Research are provided here courtesy of The Science Foundation of China Publication Department, The National Natural Science Foundation of China ACTIONS View on publisher site PDF (2.4 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top