Marine ecosystem ecology at a crossroads

Introduction

Marine ecosystem ecology links biodiversity, species interactions, trophic structure, habitat complexity, biogeochemical cycling, ecosystem functioning, ecosystem services, and human pressures across scales. Research published in the Marine Ecosystem Ecology section of Frontiers in Marine Science in recent years has reflected this integrative role, with a strong emphasis on understanding the contribution of marine biodiversity to ecosystem functioning, which lies at the core of the section’s scientific identity (Borja et al., 2020). Studies assessing ocean health have had the greatest scientific impact and have been highly cited (e.g., Borja et al., 2016; Aylagas et al., 2016; Danovaro et al., 2016), highlighting the growing concern of both the research community and society about the condition of marine ecosystems. Research examining relationships between human pressures and ecosystem responses has similarly emerged as a dominant theme (e.g., Katsanevakis et al., 2014; Patrício et al., 2016; Angiolillo and Fortibuoni, 2020; Matos et al., 2025). These research themes continue to define Marine Ecosystem Ecology, but the context in which the discipline operates has changed substantially since the grand challenges outlined by Borja et al. (2020). Many of those challenges remain unresolved, including understanding biodiversity–ecosystem functioning relationships, cumulative impacts and cause–effect pathways, predicting future ecosystem states, and translating ecological knowledge into effective conservation and restoration. What has changed is their urgency and the capacity (and increasingly the obligation) to address them.

These priorities coincide with a period of unprecedented environmental change. Climate change (Smith et al., 2023), biodiversity loss (Nikolaou and Katsanevakis, 2023), biological invasions (Haubrock et al., 2026a, b), habitat degradation (Friess et al., 2019; Turschwell et al., 2021), emerging types of pollution (Villarrubia-Gómez et al., 2018; Mitchelmore et al., 2021; Chen et al., 2025), and overexploitation (Dulvy et al., 2021) are acting together (Gissi et al., 2021), challenging the capacity of marine ecosystems to maintain their structure, functions, and services. Accelerating climate change, recurrent extreme events, and shifting species distributions are creating novel and dynamic ecological conditions, reducing the reliability of historical baselines for anticipating future ecosystem states. Simultaneously, advances in observing systems, molecular tools, autonomous technologies, artificial intelligence, and ecological modelling are fundamentally transforming our ability to monitor and understand marine ecosystems (Borja et al., 2026). However, greater observational and computational capacity does not automatically translate into greater ecological understanding, stronger scientific evidence or more effective management. Marine observations remain fragmented, geographically uneven and often dependent on short-term projects, while biological models are still constrained by limited mechanistic understanding of interactions, rates and ecological responses to multiple pressures (Sprovieri et al., 2026).

Growing demands for evidence-based conservation, ecosystem restoration, climate adaptation, and sustainable ocean governance require marine ecosystem ecology to move beyond documenting ecological change toward anticipating ecosystem trajectories and informing timely management responses. This requirement has become more immediate since 2020 as major policy commitments, including the Kunming–Montreal Global Biodiversity Framework, the Agreement on Biodiversity Beyond National Jurisdiction and, in Europe, the Nature Restoration Regulation and the European Ocean Pact, increasingly require science to inform concrete decisions on ecosystem protection, restoration and adaptive management. This need is illustrated by recent assessments of European seas, where substantial progress in documenting environmental degradation has not yet translated into corresponding ecological recovery, and where the environmental status is generally not good despite improved assessment frameworks (Nikolaou et al., 2025). This implementation–recovery gap reflects persistent cumulative pressures, incomplete and uneven implementation and enforcement of existing measures, fragmented governance across sectors and jurisdictions, misaligned financing, and ecological response times that often exceed policy cycles, with climate change further eroding potential gains (Andersen et al., 2026). Similarly, recent syntheses of estuarine and coastal research conclude that a substantial knowledge and methodological base already exists for monitoring, cumulative-impact assessment, spatial planning and restoration, but that its coordinated uptake into policy and practice remains a central challenge (Lahl et al., 2026).

The United Nations Decade of Ocean Science for Sustainable Development has reinforced this transition by emphasizing interdisciplinary and solution-oriented science that supports ecosystem protection, restoration, climate resilience, and equitable access to knowledge and technology. Marine ecosystem ecology is therefore at a crossroads. The discipline can continue largely through fragmented, project-based and retrospective approaches that document ecological patterns and change, or move towards sustained, integrated and mechanistically grounded science capable of anticipating ecosystem trajectories and supporting decisions under uncertainty. The latter does not replace fundamental ecological research; rather, credible prediction depends on stronger causal understanding of biodiversity–function relationships, interacting pressures, ecological thresholds and recovery processes. Here, prediction encompasses near-term ecological forecasts, longer-term projections under alternative environmental and management scenarios, and early-warning indicators of emerging change (e.g., Brodie et al., 2023), supported by sustained observations, mechanistic and statistical models, rigorous validation against independent observations, and explicit quantification of uncertainty. The challenge is no longer simply to generate more observations, but to integrate sustained observations with process understanding, models and decision-support systems and to translate the resulting knowledge into timely action (Sprovieri et al., 2026; Lahl et al., 2026). Without this transition, growing data volumes and technological sophistication may coexist with persistent geographic and taxonomic biases, weak attribution of ecological change, unreliable extrapolation to novel conditions, and management responses that lag behind rapidly changing ecosystems (Groom et al., 2026). Marine ecosystem ecology is therefore increasingly challenged to provide predictive, decision-ready science that enables societies to anticipate, adapt to, and steer ecological change.

Key scientific questions

The central scientific question in marine ecosystem ecology is how marine ecosystems maintain biodiversity, ecosystem functioning, and ecosystem services under environmental change and intensifying cumulative pressures, and how declining resilience can be detected before abrupt transitions occur. Here, resilience refers to the capacity of an ecosystem to persist and maintain its state and functions in the face of disturbance through two complementary components, resistance and recovery, and should be specified in relation to the ecosystem component or function considered, the relevant pressure or disturbance, and the spatial and temporal scale of assessment (Ingrisch and Bahn, 2018). Adaptation is treated here as adjustment that allows key functions to be maintained under changing conditions, whereas transformation entails reorganization into a qualitatively different ecosystem state (IPCC, 2022; Hobbs et al., 2009).

Climate change has further increased the urgency of this question, often forcing us to reconsider conservation and sustainability targets (Katsanevakis et al., 2026). Ocean warming is accelerating, while marine heatwaves have become more frequent, intense, and prolonged (Konsta et al., 2025), substantially affecting marine biota (Smale et al., 2019; Smith et al., 2023; Konsta et al., 2026). These extreme events can trigger mass mortality episodes, alter species distributions, restructure food webs, facilitate invasions, and compromise ecosystem services. Recent observations in the Mediterranean Sea demonstrate that marine heatwaves can induce recurrent mass mortality events across multiple taxa, raising concerns regarding the capacity of ecosystems to recover under repeated disturbances (Garrabou et al., 2022). Accelerating ocean warming and recurrent disturbances may progressively erode ecosystem resilience and increase the likelihood of crossing ecological tipping points, beyond which abrupt and potentially irreversible shifts in ecosystem structure and functioning can occur (Dakos et al., 2019; Lenton et al., 2025).

Understanding resilience also requires moving beyond species inventories alone. Biodiversity contributes fundamentally to ecosystem stability, productivity, carbon sequestration, fisheries production, and coastal protection (Worm et al., 2006). Yet the true magnitude of biodiversity loss remains uncertain, as many taxonomic groups, habitats, and ocean regions remain poorly studied. Marine biodiversity observations remain strongly biased towards developed regions and shallow waters. About half of the global ocean is insufficiently sampled for marine animals, with >160 million km² below 200 m lacking data and major gaps in equatorial and Global South waters (Saeedi, 2026). Long-term biological observations similarly undersample the deep ocean, while programs monitoring key coastal habitat-formers, such as seagrasses, mangroves, hard corals and macroalgae, cover only ~6% of the global coastal zone (Satterthwaite et al., 2021). Closing these gaps requires sustaining and coordinating existing programs, expanding observations in poorly sampled coasts and deep-ocean basins, integrating molecular, imaging and acoustic technologies into existing platforms, and improving data interoperability and local capacity (Satterthwaite et al., 2021; Hassoun et al., 2024). A clear example of the value of closing such gaps comes from Tara Oceans: globally coordinated taxonomic and metagenomic sampling revealed plankton lineages and interaction networks strongly associated with carbon export, linking previously undercharacterized biodiversity to the functioning of the biological carbon pump (Guidi et al., 2016). Significant knowledge gaps also persist regarding the links between biodiversity patterns, genomic diversity, species traits, and ecosystem functioning, leaving much of the ocean’s functional diversity effectively unexplored. Closing these gaps is essential for understanding how biodiversity change affects ecosystem processes, trophic interactions, ecosystem multifunctionality, and the capacity of marine ecosystems to maintain functions and services under environmental change. Identifying early-warning signals, quantifying resilience, and determining management-relevant ecological thresholds are therefore major priorities for the coming decade.

Management and policy gaps

One of the most significant impediments to effective marine stewardship is the persistent mismatch between scientific understanding and policy implementation. Cumulative-impact assessment illustrates both substantial methodological progress and this continuing implementation gap. Approaches have advanced from largely additive spatial overlays towards spatially explicit, receptor-specific and risk-based frameworks (Stelzenmüller et al., 2018) that link activities, pressures and ecosystem sensitivities and increasingly incorporate quantitative exposure–effect information and scenario analysis (Piet et al., 2023; Esquerré et al., 2026). These approaches have begun to enter formal planning practice; for example, Sweden used cumulative-impact scenario analysis to compare alternatives and support environmental assessment during development of its national marine spatial plans (Hammar et al., 2020).

Regulatory systems frequently remain sectoral, fragmented, and poorly adapted to account for interacting pressures operating across multiple spatial and temporal scales. Scientific evidence is often translated into policy through simplified indicators and targets that may fail to capture ecological complexity (Salas Herrero et al., 2019; Phillips et al., 2026). This gap is particularly evident in marine spatial planning, biodiversity conservation, and ecosystem restoration. Progress towards ecosystem-based management also requires improved understanding of cause–effect pathways linking pressures, ecological responses, ecosystem functioning, and ecosystem services, thereby enabling more robust attribution of environmental change and evaluation of management effectiveness (Long et al., 2015; Papadopoulou et al., 2025).

Area-based conservation provides a clear example. The Kunming–Montreal Global Biodiversity Framework established ambitious targets for the protection of marine and terrestrial ecosystems, including the goal of conserving at least 30% of the ocean by 2030 (CBD, 2022). Similarly, the Agreement on Biodiversity Beyond National Jurisdiction (United Nations, 2023) creates opportunities for expanding conservation efforts into the high seas. However, numerical targets alone do not guarantee biodiversity outcomes. Systematic conservation planning principles and recent analyses suggest that conservation effectiveness depends not only on coverage, but also on representation, connectivity, management quality, consideration of socio-economic trade-offs, ecological integrity, and the inclusion of climate refugia (Grorud-Colvert et al., 2021; Sala et al., 2021; Caldeira et al., 2023; Giakoumi et al., 2025). Greater emphasis should therefore be placed on evaluating the ecological effectiveness of conservation interventions, including marine protected areas and other area-based conservation measures under rapidly changing environmental conditions.

Another gap concerns equity and capacity. Marine ecological knowledge is geographically uneven: observations and research effort remain disproportionately concentrated in well-resourced regions, particularly Europe, North America and Australia, while large knowledge gaps persist across tropical and equatorial waters, especially in Africa, Southeast Asia and many Small Island Developing States (Menegotto and Rangel, 2018; Bell et al., 2023; Saeedi, 2026). These disparities partly reflect unequal access to research vessels, long-term observing systems, laboratories, analytical technologies, data infrastructures, expertise and sustained research funding; for example, just ten countries operate two-thirds of the world’s research vessels (IOC-UNESCO, 2026). Such inequalities influence not only which ecosystems and pressures are studied but also which questions are prioritized and which evidence informs management decisions. Increasing research effort in poorly studied regions without strengthening local scientific leadership can perpetuate rather than reduce these inequalities; marine research in biodiversity-rich tropical countries continues to show evidence of “parachute science” (Stefanoudis et al., 2021), in which research is led predominantly by institutions from higher-income countries with limited opportunities for local capacity development (Ahmadia et al., 2021; de Vos et al., 2023). Sustainable ocean governance therefore requires not only filling geographical knowledge gaps but also strengthening the capacity of scientists and institutions in currently underrepresented regions to generate, interpret and apply marine ecological knowledge, while ensuring more equitable participation in decision-making processes (Crosman et al., 2022; Caldeira et al., 2023).

A further governance gap is the temporal mismatch between ecological change and management cycles. Rapid ecological responses can occur faster than monitoring, policy revision, and management implementation, limiting the capacity for timely intervention (Estes et al., 2018).

Forthcoming challenges

Ecological non-stationarity is a fundamental challenge. Environmental conditions are changing so rapidly that historical baselines may no longer provide appropriate references for ecosystem assessment, conservation, and restoration. Ecosystems increasingly experience repeated disturbances, making it difficult to distinguish between degradation, adaptation, and transformation to novel ecological states (Hobbs et al., 2009; Hemraj and Carstensen, 2025). Understanding the conditions under which ecosystems can maintain key functions and services through adaptation or instead transform into novel states under future climates will require new conceptual and methodological approaches (IPCC, 2022).

A second challenge is the rapid expansion of marine data volumes. Advances in remote sensing, autonomous platforms, environmental genomics, imaging systems, acoustics, and citizen science are producing unprecedented amounts of information (Guidi et al., 2020). However, increased data availability does not automatically translate into improved ecological understanding. Data fragmentation, insufficient interoperability, inconsistent metadata standards, limited long-term stewardship, and uneven capacity to exploit increasingly data-intensive technologies may constrain the capacity to synthesize observations and generate actionable knowledge (Groom et al., 2025).

The erosion of traditional expertise presents another significant challenge. Taxonomy, systematics, and natural history remain fundamental components of marine ecology, yet many countries are experiencing declining investment in these disciplines (Wheeler et al., 2004; Engel et al., 2021). Molecular approaches and automated identification tools cannot fully substitute for taxonomic expertise, voucher collections, and field-based ecological knowledge (Hortal et al., 2015). The continued loss of taxonomic capacity may limit the interpretation of increasingly sophisticated biodiversity datasets and our ability to understand ecological processes and ecosystem change.

Finally, restoration science faces growing uncertainties. Active restoration now spans habitats from coral reefs, seagrasses and mangroves to macroalgal forests and deep-sea corals. A global synthesis of 764 interventions reported an average success of ~64%, but substantial habitat-specific variation and ~9% complete failure (Danovaro et al., 2025). Success is still assessed mainly through survival of restored habitat-formers, with biodiversity and ecosystem functioning evaluated much less often, highlighting the need for broader measures of restoration success. Scaling remains difficult: many projects are small and short-lived, costs can be high, and long-term persistence is often poorly documented (Bayraktarov et al., 2016; Fraschetti et al., 2021). Climate change further alters site suitability and may reduce the relevance of historical reference conditions (Coleman et al., 2020; Sheaves et al., 2021). Restoration should therefore target sites likely to remain environmentally suitable, include long-term monitoring, and be integrated with measures that address the pressures causing degradation; it should complement, not substitute for, protection and prevention (Eger et al., 2025). Restoration interventions can also serve as large-scale experiments for investigating resilience, recovery trajectories, ecosystem functioning, and thresholds of ecological reversibility.

Beyond scientific and technological challenges, changing geopolitical priorities and increasing competition for public investment may reduce support for long-term marine research programs, observational infrastructures, and ecosystem restoration initiatives (Sprovieri et al., 2026). Sustaining investment in marine science will therefore require stronger articulation of its societal relevance and contribution to climate resilience, food security, and human well-being.

Pathways forward

Addressing these scientific, implementation and emerging challenges requires a transition towards sustained, integrated and anticipatory marine ecosystem science. Long-term observing systems are essential for tracking ecosystem trajectories and detecting early signs of resilience loss. They should be regarded as strategic infrastructure, analogous to meteorological and Earth-observation systems, supporting not only research but also climate adaptation, ecosystem management, restoration planning, and environmental security. Emphasis should be placed on expanding biological and ecological observations, which remain considerably less developed than physical ocean monitoring despite their importance for ecosystem assessment and forecasting. Integrating field observations, remote sensing, autonomous platforms, environmental DNA, ecosystem modelling, and ecological forecasting offers opportunities to improve both scientific understanding and management preparedness. These complementary approaches allow observations to be matched to the spatial and temporal scales of ecological processes.

Agile data infrastructures are another priority. Marine ecosystem ecology increasingly depends on interoperable and FAIR (Findable, Accessible, Interoperable and Reusable) data systems that facilitate rapid knowledge generation and policy uptake. Digital Twins of the Ocean (DTOs), i.e., dynamic virtual representations that integrate observations and models to simulate ocean states and test “what-if” scenarios, could strengthen conservation, restoration and adaptive management. Their current limitations, including sparse biological data, model uncertainty, validation needs and high computational demands, can be addressed through expanded observing networks, interoperable and open data systems, improved validation and uncertainty quantification, shared computational infrastructure, and capacity development (Tzachor et al., 2023; Lin et al., 2026). Strengthening these foundations can make DTOs increasingly reliable and accessible tools for linking marine observations, prediction and decision-making.

Capacity development should also be considered a central component of scientific infrastructure, particularly in regions with persistent observational and technological gaps, including much of Africa, tropical Asia, Small Island Developing States, and parts of Latin America and the Caribbean (Menegotto and Rangel, 2018; Bell et al., 2023; Saeedi, 2026). Closing these gaps requires investments in training networks, regional centers of excellence, shared access to research and analytical infrastructure, and long-term institutional partnerships that support locally led research and equitable participation in priority setting, data ownership, and scientific publication. Such investments and stronger South–South and North–South collaborations can improve scientific equity and reduce geographic biases in global assessments of marine biodiversity and ecosystem change (Menegotto and Rangel, 2018; Saeedi, 2026).

Stronger integration between science, policy, industry, and society is equally important. Co-production approaches that involve managers, local communities, and other stakeholders can facilitate the development of indicators, thresholds, and management objectives that are scientifically robust and socially relevant. Indigenous Peoples should be recognized as rights-holders and holders of distinct knowledge systems. Their long-term place-based observations and culturally grounded indicators can complement scientific monitoring and governance, while equitable co-production should respect reciprocity and Indigenous data sovereignty over how knowledge and data are collected, used, and shared (Carroll et al., 2020; Reid et al., 2021; Baron-Aguilar et al., 2025). Ecosystem services continue to provide a valuable framework for connecting ecological processes with human well-being, food security, climate regulation, public health, and sustainable economic development.

Marine ecosystem ecology should therefore move beyond describing ecosystem patterns and processes towards actively supporting decision-making under uncertainty. This transition will require stronger interdisciplinary collaborations involving ecology, oceanography, genomics, social sciences, economics, engineering, and computer science.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. As none of the authors is native English speaker, we used ChatGPT to improve the clarity and readability of the text. The authors carefully reviewed and approved all revisions and confirm that the AI-assisted language editing did not alter the scientific content, interpretation, or views expressed in the manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Charitha Bandula Pattiaratchi, University of Western Australia, Australia

Reviewed by: Ryan Guild, University of Prince Edward Island, Canada