Underwater Legged Robots (ULRs) offer unprecedented stability and precision for seabed mobility. However, the translation of these systems to real-world deployment requires high-fidelity virtual environments for the development and validation of control and autonomy algorithms. To address this, this research presents a comprehensive review of robotic simulators from the past decade. The analysis evaluates them across hydrodynamics, seabed interaction, vehicle and sensor modeling, and Artificial Intelligence (AI) integration, while identifying critical gaps in the state of the art that hinder the effective simulation of ULRs.Complementing this review, this paper benchmarks three simulation architectures through a unified hexapod ULR case study: the general-purpose Gazebo Sim, the marine-specific Stonefish, and the AI-driven NVIDIA Isaac Sim. To enable an equitable physical comparison, a custom, GPU-accelerated hydrodynamics plugin was implemented within Isaac Sim to resolve its lack of baseline marine physics.The review reveals that current marine simulation platforms focus primarily on surface and water-column operations, failing to integrate multi-contact terrestrial mechanics with benthic hydrodynamics, required for advanced ULR locomotion. The comparative analysis demonstrates that, while framework selection depends on research objectives, successfully integrating multibody marine physics within parallelized architectures establishes the computational foundation necessary for the optimization of AI-driven bio-inspired ULR locomotion.

Organism movement is a key process in the transfer of individuals, genes, functional traits, matter, and energy among habitat patches, at sea and across the land-sea interface. The resulting fluxes, collectively termed marine functional connectivity (MFC), underpin planetary health and an array of ecosystem services. The ecological and economic impacts of rapid environmental change, including climate change, overexploitation, habitat loss and fragmentation, and the global transport of nonindigenous species make accurate estimation and prediction of MFC patterns paramount. However, estimating MFC is challenging given the relative inaccessibility of the oceans and the small size of many of the organisms and life stages with the highest dispersal potential. Here, we provide a methodological roadmap to help researchers and stakeholders understand, use, and integrate different tools to estimate organism movement and connectivity, focusing on (1) tagging and telemetry, (2) analysis of chemical markers in body tissues and structures, (3) genetics, and (4) numerical modeling. We describe method strengths and weaknesses, and the spatiotemporal resolution and scale of resulting connectivity estimates. Ancillary and emerging methods to estimate MFC are also reviewed. We then present case studies that have successfully applied or integrated different methods, particularly to support (1) marine protected area design, (2) global change predictions, focusing on climate change and bioinvasions, and (3) fisheries management. Finally, we highlight methodological innovations and concepts that promise to transform MFC research in the future.

To meet the needs of the future, marine environmental monitoring must develop methods to efficiently combine and utilise data from a diverse range of sources (e.g., satellite imagery, sensor networks, acoustic data). Generative Artificial Intelligence (GenAI) is uniquely suited to aid with this by enabling the synthesis and integration of heterogeneous and often incomplete data. Its ability to learn underlying statistical patterns supports data fusion, imputation, and enhanced interpretation across sources. GenAI also introduces novel modelling approaches to tackle ecological uncertainties and improve predictive insight. Here, we present a comprehensive overview of GenAI applications in marine ecological monitoring, emphasising its potential to improve data quality control, automate species identification, and support the creation of digital twins. We also highlight key research challenges, such as managing model bias and ensuring system transparency, and outline future directions for integrating GenAI into sustainable marine ecological monitoring and management.

Seabed observatories provide long-term, multiparametric ecological data but typically lack near-real-time communication capabilities. This limitation constrains adaptive environmental monitoring and early fault detection. To address this gap, we present a new generation of autonomous pop-up system composed of a server plus a group of buoys designed to be integrated on stand-alone benthic observatories. Each pop-up system wirelessly downloads sensor and imagery datasets from the seabed node via short-range Wi-Fi into a pop-up buoy, detaches using a low-power electro-permanent magnet release system, ascends to the surface, and transmits data through the ultra-low-bandwidth, ARGOS satellite constellation. The system includes synchronized low-power routines, data management protocols, and a dedicated JPEG reduction and segmentation workflow. Laboratory tests validated underwater Wi-Fi performance up to 40 mm separation, achieving transfer rates of up to 1.35 Mbit/s for 1 MB payloads, as well as reliable magnet release under varying buoyancy conditions. Field trials at 118 m depth within a Mediterranean Marine Protected Area provided a proof-of-concept demonstration of the complete end-to-end operation, including periodic buoy wake-ups, successful seabed data download, accurate timed release, rapid ascent, satellite telemetry, and recovery. These results demonstrate the feasibility of the proposed approach and support its future integration into deep-sea observatories as an energy-efficient and scalable solution for near-real-time ecological monitoring.

Fishery No-Take Zones (FNTZs) are increasingly being used as a spatial management tool to promote marine habitat recovery and enhance the sustainable use of fishery resources. To evaluate their effectiveness in shelf edge and upper continental slope habitats, a characterization of sessile and motile fauna within eleven FNTZs implemented along the Catalan margin was conducted. Surveys were conducted between 2023 and 2024 using Remote Operated Vehicle (ROV) video-transects across two depth ranges (100–300 m and > 300 m), covering both protected and adjacent Control areas. Sessile and motile fauna were analyzed separately, to assess the different responses in organisms with different life strategies. Results revealed that the effects of protection varied across depths and locations. Sessile fauna exhibited significantly higher densities, and diversity in several FNTZs, especially at the shelf edge, while motile fauna displayed less consistent trends. Community analyses showed differences between protected and control sites, with FNTZs hosting more habitat-forming species opposed to motile species which were more represented in Control areas. Despite initial signs of structural differentiation, results indicate that ecological recovery in soft-sediment at this depth remains slow and spatially variable, likely restricted by the habitat physical degradation consequence of decades of trawling. This research provides the first regional-scale, non-destructive assessment of FNTZs effectiveness on Mediterranean soft sediments and establishes critical reference data for future monitoring.

No-take reserves (NTRs) are marine protected areas where fishing is prohibited. They contribute to the restoration of marine habitats and communities impacted by fishing, and can provide spillover benefits to neighbouring areas. This study investigates the restoration effectiveness of the 10 km2 deep-sea NTR "Cigala de Rosas-Palamos" (350-475 m depth). It focuses on the benthic megafauna community structure and size spectra associated with the fishery-targeted stock of Norway lobster (Nephrops norvegicus) in the Northwestern Mediterranean Sea. Experimental demersal trawl surveys were conducted before the NTR's implementation in 2017, and four years later in 2021, both outside and inside the NTR. All individuals were classified and sized, to generate ecological indicators of groups similarity, diversity, density, biomass, and size spectra. All indicators improved four years after fishing halted inside the NTR. A new normalised aggregated size spectrum was developed for the full community to address differences in species groups catchabilities, which revealed an overall worsening (steepening) community size structure outside of the NTR and an improvement (shallowing) inside it. The crustacean community, the main local fishery target, had a likelihood-based size spectrum exponent b of-1.28 in 2017. By 2021 it became steeper (-1.94) outside the NTR, reflecting the loss of larger individuals under continued fishing pressure, whereas inside the NTR it became slightly shallower (-1.26), consistent with a recovery of the size structure upon protection. These findings suggest the broad recovery of demersal communities beyond commercially targeted species, highlighting the effectiveness of NTRs as tools for deep-sea ecosystems conservation and restoration.

Cabled multiparametric observatories are sustaining ecological monitoring by collecting long‐term real‐time biological and environmental data. Here, we investigated fish communities by sampling environmental DNA (eDNA) over 4 days near the multiparametric cabled video‐observatory OBSEA (Northwestern Mediterranean Sea). The multi‐marker eDNA metabarcoding approach resulted in an increased species detection and helped provide a more comprehensive view of the local fish community when combined with imaging data. These results underline the potential of omics methods in long‐term monitoring of economically and ecologically important fish species.

The implementation of Marine Protected Areas (MPAs) has increased in the last few decades, creating a deep impact on ocean governance and local fishery communities. Despite their effectiveness being actively debated, initial results show that this spatial management could promote the recovery of depleted marine communities and species, and ecosystem services such as increased fishing profits, especially in nearby areas. This could potentially maintain the economic levels of local and regional under-pressure economies while improving the overall marine biodiversity. However, the lack of ecological monitoring systems for long-lasting, non-invasive, and multiparametric measurements in near-real-time is a roadblock to national-level monitoring programs in those areas. Here we present a set of technological solutions, including robotics, biologging, and machine learning approaches to monitor MPAs, developed and validated in different deep-sea field experiments. The outcomes of these trials have yielded insights into the fine-scale behavioral ecology of different deep-sea species and the recovery dynamics of overexploited seafloors. With these tools, it is possible to monitor restored areas via ecological indicators, that are key to informing and empowering policymakers, conservation biologists, and fishery ecology experts, propelling an effective approach to the conservation and governance of marine ecosystems.

Black and white slags are common by-products of the metallurgical industry, with white slags having limited recycling applications. In this study, different circular artificial reef (AR) designs were compared using AR Ecosystem Index Transformation (AREIT) metrics and structural stability tests. The three-layer "Puzzle Reef" demonstrated the highest structural integrity and was selected for deployment, which was performed on July 18th, 2023 at the OBSEA. Although extreme current conditions during October-November 2023 caused minor displacement of components, the structure remained intact overall. The reef's influence on the local fish community was evaluated through high-frequency photo monitoring (one image per min) from July 18th, 2023, to October 24th, 2024, using an AI-automated procedure for fish classification and counting. In parallel, colonizing assemblages were analysed in the laboratory from two reef components retrieved on December 16th, 2024. The visual census recorded 18 bony fish species and one cartilaginous fish family, totalling 1,228,915 individuals (ind.). Fish biodiversity and abundance increased over time, as did AR Multimeric Indices, indicating a positive ecological trend following AR deployment. These indices, together with AREIT, are proposed as valuable tools to inform global marine habitat policies. Biofouling analysis revealed a colonizing assemblage of 2930 ind.center dot m2 across 59 species, with a total dry biomass of 23.23 mg center dot m2. The assemblage, largely dominated by balanids, crustaceans, polychaetes, gastropods, and bivalves, was mostly composed of small-sized individuals with a mean biomass of 4.6 mu g. Nevertheless, the Abundance-Biomass Comparisons curve indicated a stable and relatively mature community, characteristic of relatively low-disturbance conditions.

Human activity depends on the oceans for food, transportation, leisure, and many more purposes. Oceans cover 70% of the Earth’s surface, but most of them are unknown to humankind. This is the reason why underwater imaging is a valuable resource asset to Marine Science. Images are acquired with observing systems, e.g. autonomous underwater vehicles or underwater observatories, that presently transmit all the raw data to land stations. However, the transfer of such an amount of data could be challenging, considering the limited power supply and transmission bandwidth of these systems. In this paper, we discuss these aspects, and in particular how it is possible to couple Edge and Cloud computing for effective management of the full processing pipeline according to the Compute Continuum paradigm.

Monitoring the effects of climate change and other multi-years processes on coastal ecosystems require long-term datasets that may extend into decades. One tool to achieve this are cabled seafloor observatories that can collect continual streams of environmental and biological data as long as the equipment is maintained. Here, we used 10-years of time-lapse images (every 30 mins) from the OBSEA seafloor cabled observatory located at 20 m depth, four km offshore from Vilanova i la Geltrú (Spain) coast, to characterize temporal trends in fish community dynamics. These temporal trends were compared to in situ and remotely-sensed (MODIS-Aqua) data on temperature, salinity, and chlorophyll-a concentration (Chl-a). We observed a reduction in fish diversity over time and an increase in species turnover. Specifically, there was a decrease in the relative abundance of fish species at the lowest trophic levels alongside an increase in predators, suggesting a top-down effect. Of temperature, salinity, and Chl-a, only salinity exhibited a significant change over time. Nevertheless, the Generalized Additive Models (GAMs) revealed significant correlations between fish biodiversity indices and both temperature and Chl-a. Following models results we concluded that environmental variables affected the local fish community only at seasonal level. Including more environmental variables, such as fishing activity and pollution, in the applied models may help explain the detected decreases in biodiversity.

Effective marine conservation and management require ecological monitoring in the form of intensive real-time data collection over large spatial scales. The combined use of fixed platforms (e.g., cabled observatories) and research vessels with platforms of different levels of teleoperated autonomy (e.g., remotely operated vehicles (ROVs) and autonomous underwater vehicles (AUVs) can contribute to the acquisition of large multiparametric biological and environmental data. If those data are spatially combined, sufficient spatial coverage can be achieved for ecological monitoring. A digital twin of the ocean (DTO) approach can then be used as a virtual representation of that monitored space, enabling multiparametric analyses of environmental patterns and processes affecting biodiversity and species distributions, as well as socioeconomic activities. Here, we propose a general architecture for a DTO centred on real-time data collection from local networks on fixed and mobile platforms, such as the physical twin observers (PTO), which is synergistically merged with platforms operating at large geographic scales. We describe a roadmap to achieve this DTO via 4 key steps: (1) acquisition of in situ data with a robotic network of platforms; (2) the application of AI in image processing for extracting biological data; (3) big data management with data bubbles; and (4) development of the resulting DTO framework for providing ecosystem monitoring via the computation of ecological indicators and socioecological modelling.

Pop-up buoys present a promising solution for obtaining diverse ocean data from stand-alone, seafloor oceanographic platforms without needing to extract them to the surface. Once these pop-up buoys have stored part of the data, they ascend to the surface with the purpose of sending fragments of data via satellite. Using a Software-Defined Radio (SDR) receiver installed on a surface vehicle to capture the GPS messages is a significant step towards the automatization of geolocating the pop-up buoys and tracking their trajectory as they drift. Here we have demonstrated their feasibility and performance in laboratory condition tests and field campaigns. In the future, the implementation of an unmanned vehicle will be crucial to locate and establish direct communication with the pop-up buoy and assemble all stored data without relying on the passage of satellites. Additionally, it could be used as an aid to localise and recover these devices, reducing the time spent in the sea by the researchers and therefore their cost.

Ecological studies and restoration efforts in deep‐sea environments are a priority for international programs aimed at preserving biodiversity and recovering ocean health. These efforts, however, face technical challenges, particularly in the controlled placement of megafauna individuals near the seabed. Traditional methods, such as diving with autonomous breathing equipment, are generally impractical for large‐scale restoration beyond 50 m depth, while releasing organisms at the sea surface leads to high dispersion rates, reducing their chances of reaching the intended location. To address these challenges, we have developed and tested releaser devices equipped with enclosed compartments that mechanically open at a specified distance from the seabed, allowing for precise organism release. We designed and deployed various versions of these devices, each suited for different field applications, including active restoration actions and ecological studies requiring the safe deployment of individuals near the seabed. Technical descriptions of these devices are provided for replication and further adaptation. The devices demonstrated their effectiveness in active restoration efforts, particularly in the mass release of cold‐water corals. Additionally, they facilitated ecological studies involving the acoustic tagging of megafauna, allowing for improved monitoring of their daily and spatial behavior. The proposed animal release devices offer a practical solution for deep‐sea restoration and research, ensuring precise organism deployment at designated depths. Their design enhances efficiency in both marine conservation and scientific investigations.

Many marine monitoring infrastructures continuously collect biological and abiotic data, yet user-friendly interfaces for visualizing and translating this knowledge remain limited. This gap persists due to challenges associated with effective frameworks and tools for processing and analysing biological data in real-time. In this paper, we outline a roadmap for developing a Graphical User Interface (GUI) tailored to the continuous monitoring of ecological data with a focus on supporting marine scientists and engineers. To guide this process, we reviewed existing GUIs for visualizing marine ecological data and surveyed 43 experts to identify current gaps and key priorities. Our findings revealed a key dichotomy: GUIs either represent diverse array of biological data or perform inadequately when attempting to transfer qualitative observations to quantitative insights. Finally, we propose a stepwise roadmap for GUI development within the Digital Twin of the Ocean framework by starting from spatial mapping and progressing to data navigation. Ultimately, this study outlines critical considerations for integrating complex ecological data from diverse sources into an effective GUI. This GUI aims to support real-time monitoring, control remote monitoring for data collection, inform policymakers, and enhance public engagement in marine science.

Aim: The urgency for remote, reliable and scalable biodiversity monitoring amidst mounting human pressures on ecosystems has sparked worldwide interest in Passive Acoustic Monitoring (PAM), which can track life underwater and on land. However, we lack a unified methodology to report this sampling effort and a comprehensive overview of PAM coverage to gauge its potential as a global research and monitoring tool. To address this gap, we created the Worldwide Soundscapes project, a collaborative network and growing database comprising metadata from 416 datasets across all realms (terrestrial, marine, freshwater and subterranean). Location: Worldwide, 12,343 sites, all ecosystem types. Time Period: 1991 to present. Major Taxa Studied: All soniferous taxa. Methods: We synthesise sampling coverage across spatial, temporal and ecological scales using metadata describing sampling locations, deployment schedules, focal taxa and audio recording parameters. We explore global trends in biological, anthropogenic and geophysical sounds based on 168 selected recordings from 12 ecosystems across all realms. Results: Terrestrial sampling is spatially denser (46 sites per million square kilometre-Mkm(2)) than aquatic sampling (0.3 and 1.8 sites/Mkm(2) in oceans and fresh water) with only two subterranean datasets. Although diel and lunar cycles are well sampled across realms, only marine datasets (55%) comprehensively sample all seasons. Across the 12 ecosystems selected for exploring global acoustic trends, biological sounds showed contrasting diel patterns across ecosystems, declined with distance from the Equator, and were negatively correlated with anthropogenic sounds. Main Conclusions: PAM can inform macroecological studies as well as global conservation and phenology syntheses, but representation can be improved by expanding terrestrial taxonomic scope, sampling coverage in the high seas and subterranean ecosystems, and spatio-temporal replication in freshwater habitats. Overall, this worldwide PAM network holds promise to support cross-realm biodiversity research and monitoring efforts.