This study explores the role of subsurface thermal stratification in shaping Mediterranean climate dynamics over the past century. A nonlinear regression model is applied to historical temperature-depth time series (1900–2019) from the SeaDataNet archive, focusing on four key subregions: the Aegean Sea, the Gulf of Lion, the Strait of Gibraltar, and the Strait of Sicily. The analysis employs the Plateau Followed by One Phase Decay (PFOPD) model to extract depth-resolved thermal metrics, including thermocline depth, surface and deep-water temperatures, and decay rates.The results indicate a long-term pattern of summer shoaling and winter deepening of the thermocline, coupled with regional variability and basin-wide subsurface warming. The PFOPD model demonstrates high explanatory power (R2 > 0.95 in most summer time series), enabling the detection of robust climatic signals from fragmented historical data. Stratification metrics show strong covariation with atmospheric circulation regimes, highlighting their diagnostic value for climate variability assessments and early-warning systems. These findings are further contextualized by the 2024 flood event in Valencia, presented as an illustrative example of how persistent marine stratification may modulate atmospheric instability, although a formal attribution of the event would require dedicated coupled modeling.Beyond methodological innovation, the study proposes a reconceptualization of the Mediterranean Sea as a depth-integrated climatic system, where subsurface processes shape regional climate feedbacks. By bridging historical ocean data with present-day environmental challenges, this research supports improved monitoring strategies and informs climate-sensitive marine and coastal planning under ongoing warming and intensifying stratification.
Despite substantial variations in the physicochemical and biological environments around the subsurface chlorophyll maximum (SCM), how photosynthetic parameters vary around the SCM has been scarcely studied. To reveal the fine-scale profiles of photosynthetic parameters and their controlling mechanisms, seawater samples collected at approximately 3-m vertical intervals in the eastern Indian Ocean were analyzed using single turnover active fluorometry (STAF), together with other environmental parameters. The saturation irradiance of photosynthesis was higher than the in situ average irradiance of photosynthetically active radiation during the daytime throughout the sampled layers, indicating that the subsurface phytoplankton communities were under suboptimal light conditions. Throughout the studied depths, the initial slope of the photosystem II photochemical flux − irradiance curve (α-value) and maximum quantum yield of photosystem II (Fv/Fm) increased with depth, while non-photochemical quenching (NPQNSV) decreased, suggesting relief from photoinhibition and nutrient stress with depth. In addition to these general trends, at the three stations between the equator and 21°S, local minima of α and Fv/Fm values were observed around the SCM, along with increased NPQNSV, suggesting the enhanced physiological stress on photosynthesis by phytoplankton communities around the SCM in the eastern region of the South Indian Subtropical Gyre. These local variations were significantly correlated with the chlorophyll concentration and could not be explained by changes in solar irradiance or macronutrient availability. In addition to changes in phytoplankton community composition, one possible cause is iron deficiency under light limitation, which is sometimes observed in stratified subtropical waters. However, further examination in this area is required. These local variations in photosynthetic parameters can result in lower carbon fixation rates around the SCM in the eastern Indian Ocean compared to estimates based on relief from light and macronutrient limitations.
Satellite observations reveal that mesoscale eddies in the southeastern Indian Ocean (SEIO) are highly active and exhibit pronounced spatial variability in their surface characteristics. By combining satellite and Argo float observations, this study investigates the spatial variations in eddy vertical structure and volume-integrated energy in the SEIO, with particular emphasis on their relationship to eddy lifespan. Cyclonic eddies (CEs) display strong regional contrasts, whereas anticyclonic eddies (AEs) vary less. Subsurface-intensified eddies, predominantly cyclonic, feature a density core near 750 m and maximum velocity at ∼ 250 m depth, and mainly originate from the Leeuwin Current System (LCS) and the open ocean south of 30°S. Eddy lifespan is strongly correlated with volume-integrated energy, which is mainly controlled by eddy amplitude and eddy vertical structure . To further interpret this relationship, we examined energy dissipation terms in the eddy energy equations. The eddy energy dissipation term is expressed as a function of eddy structure, allowing us to quantify regional differences in energy decay rates among eddies with distinct vertical structures. In subsurface-intensified eddies, dissipation due to vertical mixing is much weaker than in surface-intensified eddies. Vertical mixing dissipates eddy potential energy more rapidly than eddy kinetic energy, while horizontal mixing shows the opposite tendency. Overall, CEs in the LCS tend to sustain longer lifespans owing to their higher volume-integrated energy and weaker vertical mixing. These findings underscore the pivotal roles of eddy structure and energy in regulating eddy lifespan.
Mesoscale eddies are prevalent in the Philippine Sea and frequently impinge on the Kuroshio Current, affecting internal tides (ITs) originating from the Luzon Strait (LS). This process is investigated in this study using idealized experiments, which provide a controlled environment to quantify variations of the M2 ITs as a single cyclonic eddy (CE) or anticyclonic eddy (AE) approaches the LS from the east. Our results reveal an asymmetric influence: AEs induce more pronounced changes in the M2 tidal conversion than CEs. Variations in conversion rates are primarily controlled by changes in bottom pressure perturbations, which are linked to the product of subtidal buoyancy gradients and IT vertical velocity. Another noteworthy finding is that these mesoscale eddies influence the M2 ITs differently in two stages: during propagation across the Philippine Sea and upon impingement with the Kuroshio Current. In the first stage, mesoscale eddies modulate the propagation of the M2 ITs through direct interaction, inducing incoherent M2 ITs in the Philippine Sea. The percentage of incoherent ITs exceeds 50% along the eddy paths. Advection from eddy-induced currents contribute significantly more than background stratification and relative vorticity. During the second stage, as the eddies impinge on the Kuroshio Current and move northward, variations in the M2 IT energetics within the LS are evident. The IT incoherence increases not only in the Philippine Sea but also in the LS and South China Sea. These findings are important for understanding the IT variability and incoherence in the Philippine Sea and South China Sea.
Mid-Trophic Level (MTL) organisms — including krill, forage fish, and mesopelagic fish — are abundant in the California Current System (CCS) and play an essential role in transferring energy and biomass from primary producers to top predators. However, their spatiotemporal distribution and variability remain poorly understood, particularly with respect to vertical structure across epipelagic and mesopelagic habitats and coastal-offshore gradients. This lack of understanding emerges from both the complexity of MTL interactions with a heterogeneous environment and the challenges associated with sampling these organisms at high spatial and temporal resolution. To address this gap, we analyze 11 years of fisheries acoustic observations in the CCS (2006–2016) to characterize the spatiotemporal dynamics of MTLs as inferred from acoustic backscatter. Acoustic observations at 38 and 120 kHz, collected during day and night across depth strata from 15 to 495 m, reveal consistent cross-shore, seasonal, and latitudinal patterns in the backscatter of acoustically defined zooplankton, epipelagic fish, and mesopelagic fish communities. These patterns include: (1) weaker cross-shore gradients in mesopelagic relative to epipelagic communities; (2) a temporal succession among communities associated with seasonal upwelling; and (3) a multimodal latitudinal distribution with distinct coastal backscatter peaks. Alongshore variability is further modulated by El Niño and the Pacific Decadal Oscillation. We investigate relationships between acoustic backscatter and co-located environmental variables from in situ, remote sensing, and reanalysis products to elucidate plausible mechanisms underlying MTL dynamics.
Isoscapes – spatial models of stable isotope ratios – are powerful tools for understanding biogeochemical cycles, food web structure, and ecological connectivity. In marine systems, however, broad-scale isoscapes are difficult to constrain because reference datasets are typically heterogeneous and opportunistic. Bayesian hierarchical models implemented through Integrated Nested Laplace Approximation (INLA) provide an efficient framework for modeling spatial patterns while accounting for non-spatial variance. We tested whether broad-scale geographic baseline isoscapes could be constrained for the Mediterranean Sea using INLA models applied to an opportunistically compiled dataset from this semi-enclosed basin, where strong environmental gradients are expected but spatially explicit baselines remain lacking. We assembled the most comprehensive dataset of δ13C and δ15N values to date for low-trophic-level, primarily epipelagic food web compartments (> 6,000 samples from > 600 sites). Although discrete regions of relatively high and low δ values were identified, particularly for δ15N, semivariograms and model outputs indicated only short-range spatial autocorrelation (ca. 40 km for δ13C; 65 km for δ15N) within this large dataset. Environmental covariates improved model fit for δ15N values but did not enhance predictive performance or alter spatial predictions. Our results demonstrate that even advanced spatial models may not compensate where sampling design is structurally mismatched relative to unknown but structured spatial and non-spatial sources of isotopic variation. Coordinated sampling designed to capture spatial variability while accounting for major sources of non-spatial variance is required to develop robust marine isoscapes supporting ecological and management applications in the Mediterranean Sea and similar regions.
Seasonal oxygen depletion has been increasingly reported on the East China Sea (ECS) shelf since the early 2000s; however, its long-term evolution and underlying drivers remain poorly understood. Using historical observations from the World Ocean Database (1951–2006), we reconstruct basin-scale variability of dissolved oxygen (DO) across the ECS shelf and analyze both seasonal patterns and multi-decadal trends. Our results show that oxygen depletion (DO ≤ 4 mg L⁻1) on the mid–outer ECS shelf has occurred regularly since the 1950s, substantially earlier than previously reported observations, which mainly documented its occurrence after the early 2000s. Oxygen depletion recurs from summer to autumn, with oxygen stress shifting from the inner shelf in late summer toward the mid–outer shelf along the 50–100 m isobaths in autumn. A significant decline in bottom DO is detected over the mid–outer shelf during 1960–1999 and is accompanied by increasing apparent oxygen utilization (AOU). By combining AOU diagnostics with a three-endmember mixing model, oxygen deficits are partitioned into locally generated consumption and remotely advected components. The results indicate that local remineralization dominates oxygen consumption, while the intrusion of oxygen-poor Kuroshio Subsurface Water reduces the background oxygen inventory and modulates the intensity of seasonal depletion. These findings reveal that oxygen depletion on the ECS shelf has a longer history and broader spatial extent than previously recognized.
Small islands are highly vulnerable to climate change as rising sea levels, warming temperatures and ocean acidification threaten marine ecosystems, livelihoods and traditional way of life. The Australian Torres Strait Islands are a low-lying archipelago between northern Australia and Papua New Guinea and face existential threats akin to small island developing states in the Pacific. We analyse trends in ocean and atmospheric variables and synthesise changes that occurred in key Torres Strait marine habitats and resources since the 1970 s. We find increasing trends in temperature, marine heatwaves, sea surface height and dry-season wind run. Other variables including solar exposure, rainfall, river discharge and net primary productivity show variability but no trend, and are typically aligned with large El Niño Southern Oscillation (ENSO) events. Habitats such as mangroves, seagrass meadows and coral reefs are naturally variable, but sensitive to compounding climate extremes, including ENSO-related anomalies and sustained ocean warming. Our study further explores the dependence of commercial and cultural fisheries on these foundation habitats, highlighting that habitat fluctuations combined with ocean warming are already impacting some fisheries (e.g. tropical rock lobster). Cultural totemic species such as dugongs and green turtle are also under pressure from climate change. Our synthesis informs efforts to respond and adapt to climate change. Moreover, we highlight that several interconnected initiatives are needed, including long-term monitoring, climate change communication, adaptive strategies for fisheries and communities, as well as strong governance and policy support alongside continued efforts to empower communities through capacity building, education, and equitable engagement.
Phytoplankton functional groups distribution and dynamics builds nutrients and carbon energy pathways in the ocean. They respond fast to changes in their environment, and the understanding of their dynamics relies on regular sampling to resolve daily to weekly scales. The Bonifacio Cyclonic Gyre (BCG) is an area prone to intermittent phytoplankton blooms triggered by westerly winds, and close to coastal areas capable of fuelling open waters with nutrients from flooding. To study phytoplankton evolution in this dynamical area, the distribution of phytoplankton functional groups in the surface waters of the western Mediterranean Sea was investigated semi-continuously (30 min) using an automated CytoSense flow cytometer coupled to a FerryBox, onboard the ferry Le Carthage, between October 20, 2016 and January 12, 2017 and along the route from Tunis (Tunisia) to Genova (Italy). The BCG signature was identified by its surface temperature anomaly and evidenced significant higher abundances of the RedPicoProk, the RedPico, the HsNano and the OraNano over the full sampling period, except for the RedNano1, the RedNano2 and the OraPicoProk. Their estimated chlorophyll a concentration per group were higher within the gyre than outside, but did not affect the carbon/chlorophyll a ratio. A high carbon/chlorophyll a ratio can indicate recently uplifted cells that are not yet adapted to high light conditions. In this case, the observed ratios suggest that the cells were not recently upwelled, but instead had time to adjust their photophysiology to surface conditions. During 19–20 December, extreme overflow occurred over Corsica and Sardinia, considered as one of the highest rainfall events of the past 20 years. The runoff was evidenced by the low salinity intrusion that lasted several days after the storm. After the runoff and within the Bonifacio gyre region, biomasses of all groups increased and remained high until the end of the sampling, except for the RedPico. The runoff triggered an early bloom end of December that persisted for several weeks, being trapped and repleted by the uplift of the isopycnals from the BCG. Regular observation of the sea surface temperature and salinity together with phytoplankton community structure provides essential insights into the fast response of phytoplankton communities to an extreme event at unprecedented resolution.
Light in the ocean is an important driver of organism behavior and primary productivity. Bioluminescence, or light produced by living organisms, contributes to marine lightscapes. While this trait is ubiquitous throughout marine species, observations of bioluminescence remain relatively sparse compared to other oceanographic variables. Here we present glider based observations of bioluminescence and atmospheric light in the north west Mediterranean Sea as a part of the BIOLUminescence Marine, Observations spatio-temporelles in situ par Planeur Sous-marin (BIOLUMOPS) campaign. We consider the relationship between these variables down to a depth of 600 m using data from two deployments in 2025. We calculate the depth at which the light field is dominated by bioluminescence in this region, termed the bioluminescence transition depth. Our findings suggest a strong diel signal in transition depths from the surface at midnight to depth ranges between 246 and 465 m at midday. This depth range approximately follows a range of photosynthetically active radiation (PAR) isolumes from 2.1 × 10−5 to 1.8 × 10−4 μmol photon m−2 s−1, consistent across both the winter and summertime deployments. These findings present a new perspective on lightscapes in the Mediterranean Sea, with implications on light-mediated behavior and the effectiveness of communication between organisms.
Zooplankton community structure has traditionally been studied using taxon-based methods, which relate species abundance and diversity to environmental variability. Although size-based approaches provide a functional perspective on ecosystem structure, their application across contrasting regions and extended vertical gradients remains limited. Here, we provide the first integrated assessment of mesozooplankton taxonomic and size-based structure by investigating the geographical, vertical, and diel distribution of biomass, abundance, diversity, and normalized biomass size spectra (NBSS) down to 700 m in five zones (Balearic Islands, Alboran Sea, Cadiz, Lisbon, and Galicia) during autumn 2020. Marked abundance and biomass differences were observed around the Iberian Peninsula, with significant spatial and depth-related variability but no consistent diel variations. NBSS slopes – used as an indicator of zooplankton community structure – were consistently flatter than −1, suggesting relatively high trophic transfer efficiency. We identified a marked shift in the NBSS slopes and size-diversity at the epipelagic – upper-mesopelagic boundary layers (∼300–500 m), with the magnitude of this transition varying based on regional productivity. In low-productivity areas (Balears and Cadiz), the transition toward larger size organisms occurred at greater depths, whereas in high-productivity regions (Lisbon and Galicia), the transition occurred shallower, hosting larger mesozooplankton closer to the surface. Finally, size diversity decreased from the southwestern Mediterranean to the North Atlantic, while abundance and biomass peaked off Lisbon, the most productive region. By linking vertical size-structure transitions to productivity gradients across a major biogeographic boundary, this study underscores the value of integrating size-based metrics into regional ecosystem assessments.
Coral reefs are experiencing widespread decline from climate change and anthropogenic stressors, increasing the need to understand the processes that maintain and replenish coral populations. Connectivity—the exchange of individuals between coral populations—underpins population recovery, yet major uncertainties and natural variability remain in larval biology and behaviour during dispersal and settlement. This review synthesises empirical and theoretical understanding of the physical and biological drivers of coral connectivity from laboratory and field studies. Rather than reviewing connectivity modelling frameworks, we focus on the underlying processes that govern connectivity in nature and that constrain and inform predictive approaches. We highlight well–studied processes such as spawning timing and identify key data gaps, including buoyancy, mortality, and competency during dispersal. High variability across taxa and locations complicates the identification of general connectivity patterns, particularly for processes that are difficult to observe in situ. Emerging techniques—including larval colouring, time–series competency assays, acoustic enrichment, AI–based behavioural tracking and genetic barcoding—offer promising avenues to address these gaps. By synthesising empirical evidence and recent methodological advances, this review highlights the processes that underpin coral connectivity and provides a basis for improving empirical measurements and informing future connectivity model development to support understanding and future applications in conservation and restoration.
The identification of the factors driving the biodiversity and metabolism of microbial components in deep-sea Antarctic ecosystems is crucial in light of their rapid transformations. Here, through a replicated and hierarchical sampling strategy, we investigated two deep-sea benthic areas in the Ross Sea (Antarctica) characterized by different bottom temperatures (ΔT ca. 1.3°C) and trophic conditions. The warmer and more oligotrophic deep seafloor showed higher bacterial and archaeal diversity and faster organic matter cycling, whilst the colder mesotrophic system displayed a higher organic matter content and microbial biomass. Proteobacterial assemblages dominated both areas, yet only 9.2% of the taxa were shared between the two sampling areas, indicating a major turnover in microbial biodiversity. Multiple linear regression models identified temperatures and organic matter as key drivers of prokaryotic assemblages and ecosystem functioning, with viral infections playing a potential role in organic matter cycling. These findings allow understanding how environmental changes in Antarctic benthic ecosystems, such as those potentially induced by climate change, could alter the biodiversity and metabolism of the microbial components and the organic matter cycling.
Mesoscale eddies are rotating vortices of water that perturb the local physical, chemical, and biological environment. In the Southern Ocean, penguins inhabit regions characterised by intense eddy activity, which can offer foraging opportunities. To better understand relationships between penguins and eddies, we collated tracking data for five species (emperor, king, chinstrap, Adélie, and macaroni penguins), totalling 3189 individuals from 59 colonies. Data were subset by colony and breeding stage to create 74 case studies. We then fitted Hidden Markov Models to penguin tracks to identify Area-Restricted-Search (ARS) behaviour (a proxy for foraging) and used Generalised Additive Mixed Models to relate behaviour to the presence of eddies. In 28 case studies, penguins displayed a pronounced increase in ARS within specific parts of eddies. All five species exhibited an association between ARS and eddies in several case studies, though substantial regional differences in association strength were observed. In life history periods when penguins experienced central place constraints, ARS was more frequently associated with eddies. By studying five extensive case studies in greater detail, we found that eddies possibly influenced trip trajectories by aggregating prey at submesoscale filaments around their peripheries, by interacting with the distribution of sea ice, and potentially by transporting key prey species in their interiors. In these case studies, eddy maturity, amplitude, and intensity also differed between eddies collocated with ARS and the background eddy field. When eddy abundance varied, foraging trip durations also varied, though with opposing trends depending on the colony and breeding stage. Eddy activity is projected to increase in the Southern Ocean, which in isolation would have mixed impacts on penguins, though the negative impacts of broad-scale changes in prey availability and sea ice concentration are likely to outweigh any impacts of changing eddy fields.
Understanding the drivers of assemblage structure across broad depth gradients remains a central challenge in abyssal and hadal ecology (3000–11000 m), particularly in disentangling the relative roles of depth and environmental heterogeneity. Scavenging amphipods are a model group for investigating drivers of deep-sea species distributions, owing to their high abundance and ubiquity. Here, we resolve amphipod assemblage structure and its drivers across a 5000 m depth gradient within the Nova Canton Trough (central Pacific Ocean) using a high-resolution dataset composed of 102 baited lander deployments spaced at ∼ 100 m depth intervals and integrated with multiscale environmental descriptors. A total of 139,271 amphipods were recovered, comprising 26 species identified using integrative taxonomy. Hierarchical clustering organised species into seven distinct assemblages spanning lower bathyal to hadal depths. Assemblages in the abyssal-hadal transition zone exhibited particularly high turnover, consistent with an ecotone nested within a broader depth-related environmental ecocline. Lower-hadal assemblages (>7000 m) were comparably cohesive, dominated by three species (Bathycallisoma schellenbergi, Hirondellea dubia, and Profundum solomoni). Multivariate analyses identified depth as the dominant gradient structuring assemblage composition, with additional contributions from in-situ temperature, regional topography, local seafloor composition, and spatial gradients. Together, these predictors explained 53.5% of total variance. Variation partitioning revealed that a substantial proportion of this variation was shared between depth and environmental predictors, reflecting covariance along the depth gradient. Despite this, environmental predictors explained an independent fraction of variation. This study represents one of the most detailed assessments of scavenging amphipod assemblages within a single hadal feature. By integrating integrative taxonomy, high-resolution sampling across a broad bathymetric gradient, and multiscale environmental variables within a seascape ecology framework, it provides new insights into the relative roles of depth, regional context, and macro- to microscale environmental heterogeneity.
Oceanic archipelagos host unique biodiversity shaped by strong geomorphological and oceanographic gradients. The Canary Islands and their adjacent seamounts is a biodiversity “crossroad” between the Atlantic and Mediterranean biogeographic zones, spanning ca. 1,300 km from the western to the eastern islands. Here, we took advantage of this geomorphological configuration to assess whether the distribution patterns of black corals (Anthozoa, Antipatharia) varied across island groups (i.e. western, central, eastern, and seamounts) in relation to key environmental and geomorphological drivers. Occurrences of antipatharians were compiled from SCUBA surveys, towed cameras and ROV deployments conducted across 144 sites and depths from 1 to ca. 1,000 m. Over 2,400 presence records were obtained, including 11 taxa from 10 genera. Black coral assemblages showed clear bathymetric partitioning, with some taxa mainly restricted to mesophotic depths, others occurring predominantly in bathyal waters, and Stichopathes spp. and Leiopathes sp. spanning broader depth ranges. Depth and substrate type were the main predictors of the distribution of most common taxa (Antipathella wollastoni, Antipathes furcata, Parantipathes spp. and Stichopathes spp.), while horizontal current speed, vertical velocity, slope and island group also influenced their occurrences. A. wollastoni and A. furcata were strongly associated with hard substrates at mesophotic depths, whereas Parantipathes spp. was more frequent in bathyal habitats and positively related to horizontal current speed. These findings provide predictive insights into the distribution patterns of black coral taxa in relation to diverse environmental and geographical factors, revealing the significance of this region as an important global hotspot, and providing essential baseline information for the sustainable management and protection of mesophotic and deep-sea coral-dominated ecosystems.
Small pelagic fish are key components of productive coastal ecosystems, yet their migration ecology remains poorly understood, causing challenges for management. We applied stable carbon and nitrogen isotope (δ13C and δ15N) analyses of eye lenses to investigate movements of European anchovy (Engraulis encrasicolus) and sardine (Sardina pilchardus) around the Iberian Peninsula. Muscle isotopes showed strong spatial heterogeneity, largely consistent between species and reflecting differences in baseline values. Eye lens centres of small anchovy, and to a lesser extent sardine, showed clear geographic variation: higher δ15N off the Atlantic south coast, lowest δ15N in the Alboran Sea, and lower δ13C off the west coast. These patterns persisted across years and fish sizes in anchovy, with only minor outliers, suggesting limited cross regional migration. An exception was the overlap between west coast and Cantabrian Sea values, consistent with connectivity supported by cohort tracking. In contrast, sardine isotopes from the west and south coasts converged into a unimodal distribution with growth, indicating frequent exchange between these regions. These findings support recent revision of stock limits that separate south and west coast anchovy stocks, but they question the current assumption of separation between western and northern Iberian anchovy stocks. Eye lens isotopes provide a powerful complementary tool to resolve nursery origin and connectivity, offering new opportunities for fisheries management in shorter time scales than molecular techniques, which is paramount to be able to cope with rapid changes of fish distribution due to climate change, and for spatially explicit management.
