The field experiment was conducted in Bazaruto Archipelago National Park, where the largest known dugong subpopulation in the Western Indian Ocean occurs. Dugongs are large, mostly herbivorous marine mammals whose populations have declined throughout their range, and visual or aerial surveys can be difficult where animals are scarce, solitary, cryptic or avoid human activity. A positive eDNA result therefore offers a non-invasive way to supplement field observations, although the signal may reflect an animal that was present nearby or recently passed through the sampled area rather than one present at the exact sampling point and time.

During the campaign, the research team followed 12 non-linear boat tracks at an average speed of about 1.5 knots through known areas of dugong occurrence. The sampling design compared surface water with near-bottom water, filtered volumes of 1.5, 30 and 60 litres, and approximately 25 millilitres of sediment. Two filter configurations were used: 0.45-micrometre polyethersulfone Waterra cartridges for the 30- and 60-litre samples and 0.22-micrometre Millipore Sterivex filters for the 1.5-litre samples. Near-bottom water was collected about one metre above the seabed with a pump mounted on a manta tow frame, while surface samples were collected from the boat or from composite samples taken at the beginning and end of each track.

The investigators developed a species-specific qPCR assay, named Ddug01, targeting a 115-base-pair segment of the mitochondrial 16S ribosomal RNA gene. In silico analysis indicated amplification specific to dugongs and the extinct Steller’s sea cow. Testing with serial dilutions of dugong tissue DNA produced a linear response from undiluted DNA through a 10^-4 dilution, with a slope of -4.819, an R² of 0.984 and an amplification efficiency of 61.3%. The lowest concentration detected in all four technical replicates was 0.944 picograms of DNA per reaction, which was assigned as both the limit of detection and limit of quantification. Amplification became inconsistent at the next tenfold dilution and was absent at lower concentrations.

Each environmental sample was amplified in eight technical replicates, and a beta-binomial mixed-effects model accounted for positive versus negative replicates while including sampling batch as a random effect. Laboratory processing was associated with markedly stronger qPCR detection than field processing, with a significant treatment effect of χ² = 15.06 and p < 0.001. The comparison concerned complete operational workflows rather than extraction location alone: field samples were processed with smaller amounts of preservation buffer, lower centrifugation capacity and, for Sterivex filters, a different extraction kit. Positive results from field processing showed that the method functioned, but the lower yield indicates a practical limitation of the field protocol.

Filtration volume and vertical position interacted significantly, with χ² = 3.875 and p = 0.049. At the low volume of 1.5 litres, near-bottom samples had about 5.8 times higher odds of producing a positive qPCR replicate than surface samples, a statistically significant difference with p = 0.013. At 60 litres, however, detection was similar between the two vertical positions. The corresponding pattern appeared when volume was compared at each depth: increasing the filtered volume from 1.5 to 60 litres significantly improved detection in surface water, with an odds ratio of 5.49 and p = 0.012, whereas it did not significantly change detection near the bottom, with p = 0.917.

The highest sample-level detection rate came from 60-litre near-bottom samples processed in the laboratory, with 9 of 11 samples positive. The researchers also detected dugong DNA in samples collected 1 hour 40 minutes and 2 hours after the last visual sightings, and at two sites where no animals had been directly observed. One event without a visual sighting produced no eDNA detection, while a known feeding site with visible feeding trails also yielded no positive signal. These outcomes illustrate the complementary nature and limitations of the two approaches: eDNA may reveal animals missed by visual surveys, but absence of a detection does not necessarily establish that dugongs were absent, particularly where tides dilute signals or DNA disperses quickly.

Sediment did not outperform the water-based methods in this pilot. None of the 10 sediment samples produced a positive result, whereas every water-sampling and extraction combination produced at least one detection. The comparison is incomplete because sediment was processed only in the field, where overall detection was lower, and the extraction kit may have performed differently with sediment than with water. The result therefore does not establish that sediment is inherently unsuitable for dugong eDNA analysis. It shows only that the tested sediment workflow offered no advantage under the conditions encountered, while sediment eDNA may integrate signals over longer periods than water samples and could provide information about past rather than recent presence.

High-volume filtration also carried practical benefits. The 0.45-micrometre Waterra cartridge filtered water at an average rate of 25 seconds per litre, allowing a 60-litre sample to be processed in less than 25 minutes at the surface and underwater without clogging. Placing the cartridge directly on the pre-pump side of the system also reduced the amount of single-use tubing consumed, a relevant advantage for large monitoring programs. The field protocol used centrifugation limited to 5,000 g and could not process the full preservation-buffer volume processed in the laboratory. Such conditions are less effective for recovering free or weakly adsorbed DNA than higher-force centrifugation or precipitation methods involving reagents that are difficult to transport because of their toxicity and volatility.

The authors therefore favor mobile surface sampling with high-volume filtration as a practical approach for large-scale dugong monitoring, while acknowledging that the proof-of-concept study included only 12 sampling events and a limited number of positive detections. Sampling volume was also partly confounded with filter configuration and, in some laboratory comparisons, extraction kit, so the results cannot be attributed exclusively to volume. The study supports further testing across broader spatial, temporal and hydrodynamic conditions, including tidal cycles, and does not directly quantify DNA persistence or transport. Even with those limitations, the work shows that carefully chosen water sampling and processing conditions can improve the recovery of rare marine-mammal eDNA and could complement conventional surveys in tropical coastal habitats.

Subject of Research: Optimization of environmental DNA sampling and extraction for detecting rare dugongs in tropical coastal waters

Article Title: Optimizing Marine eDNA Protocols for Rare Megafauna: A Case Study on Dugongs in the Western Indian Ocean

Article References: Bernagou, L., Serval, A., Beudard, F., Mathey, J., Mercky, Y., Le Bozec, C., Trotzuk, E., Paunde, K., Garcia, M., Bellemain, E., & Corse, E. (2026). Optimizing Marine eDNA Protocols for Rare Megafauna: A Case Study on Dugongs in the Western Indian Ocean. Ecology and Evolution, 16(10), Article e74500. https://doi.org/10.1002/ece3.74500

Image Credits: AI Generated

DOI: 10.1002/ece3.74500

Keywords: dugong eDNA, environmental DNA, marine monitoring, qPCR, Bazaruto Archipelago, water filtration, molecular detection, Mozambique, sea mammal conservation, Optimizing, Marine, eDNA

The research, published in the journal BioRisk, set out to carry out an initial screening of trace metal bioaccumulation in the soft tissues of wedge clams collected from different localities of the Bulgarian Black Sea coastal area, and to evaluate whether the species could serve as a bioindicator of metal pollution. Trace metals constitute a significant proportion of the pollutants in the Black Sea, where the highest concentrations are typically locked in marine sediments from which they can be released back into the water column. Because metals are persistent and bioaccumulative, their discharge into the Black Sea can have profound effects on the entire marine environment, and unlike many other pollutants they remain in soils and waters for hundreds to thousands of years.

Bivalves are particularly well suited to pollution monitoring because they are filter feeders and primary consumers in the Black Sea food chain, accumulating metals and other contaminants in their tissues and shells. Mussels of the species Mytilus galloprovincialis have traditionally been used as biomonitors of Black Sea metal pollution, but data on the spatial and temporal distribution of metal accumulation in benthic clam species from the Bulgarian coast has been quite scarce. Although the spatial variation of metal concentrations in sediments of the Bulgarian Black Sea coastal areas is well documented, no comprehensive studies on bioaccumulation of metal elements in clams from these habitats had been carried out before, despite the growing economic importance of clams for Bulgaria.

To fill this gap, the researchers collected wedge clams measuring 23 to 35 millimetres in length, either manually or from commercial providers, from sublittoral sandy habitats at representative localities along the Bulgarian coast. Sampling took place in June and September 2020, before and after the intensive touristic season, allowing the team to compare conditions on either side of the summer influx of visitors. The soft tissues of individual clams were digested with nitric acid and then analysed by Inductively Coupled Plasma Optical Emission Spectrometry, a technique known as ICP-OES, for the determination of seven elements: cadmium, chromium, copper, iron, nickel, lead and zinc. The measurements were carried out in the certified chemical laboratory at the Department of Chemistry of the Medical University-Varna, and differences between sites were tested statistically while patterns of similarity were explored with cluster analysis.

The results revealed that the content of trace metals in the wedge clams differed significantly amongst localities. Higher metal loads were present in clams from Sveti Vlas, Shkorpilovtsi, Slanchev Bryag, Ahtopol and Kranevo, several of them popular resort beaches. The highest values of lead, at 2.51 milligrams per kilogram wet weight, and cadmium, at 0.32 milligrams per kilogram, were found in samples from Sveti Vlas, together with high concentrations of chromium and copper. The highest concentration of copper, 34.12 milligrams per kilogram, along with the peak values of iron at 269.52 milligrams per kilogram and nickel at 0.32 milligrams per kilogram, were detected in wedge clams from Shkorpilovtsi. Maximum chromium content, 0.58 milligrams per kilogram, was present in samples from Slanchev Bryag, accompanied by high iron values, while the highest zinc concentration, 18.04 milligrams per kilogram, together with elevated chromium and iron, was measured in clams from Irakli. Varna Bay clams, meanwhile, recorded the highest June zinc value of 15.98 milligrams per kilogram among the tabulated sites.

Cluster analysis using Euclidean distance as a measure grouped the localities into two main clusters based on the similarity of metal concentrations in the clams. The first cluster comprised localities in the northern coastal area, namely Shkorpilovtsi, Sveti Vlas, Kranevo and Slanchev Bryag, with Shkorpilovtsi and Sveti Vlas showing the highest similarity, evidently due to the high accumulated levels of lead, cadmium and copper in clams sampled there. The second cluster included Primorsko, Arkutino, Sozopol and Ahtopol from the southern coastal area, with Varna Bay grouping together with Ahtopol because of the similarity of metal concentrations between them. As a whole, the levels of metal bioaccumulation in wedge clams from the northern localities were significantly higher than in clams from the southern localities, indicating different metal pollution levels in the northern and southern Bulgarian Black Sea coastal areas. Amongst the southern sites, Primorsko and Arkutino were most similar, and Ahtopol stood out as an exception, showing relatively higher values of copper, cadmium and nickel that linked it to Varna Bay instead.

The seasonal comparison, based on four localities sampled in both June and September 2020, found little variation overall. In wedge clams from Kranevo, however, the concentrations of copper, chromium and iron were significantly lower in September samples compared to June, and the content of chromium in clams from Varna Bay and Primorsko was also significantly lower in September. The researchers suggest that the observed seasonal variations were probably connected with local changes in the hydrological parameters of the ecosystems rather than with tourism alone, although the sampling design, spanning the period before and after the intensive touristic season, was chosen precisely to capture such pressures.

The study’s conclusions carry a double significance. On one hand, the wedge clams from localities known to have higher coastal inflows and touristic pressures, namely Varna Bay, Shkorpilovtsi, Sveti Vlas, Slanchev Bryag and Ahtopol, accumulated significantly higher metal elements in their tissues, confirming that the species registers the human footprint along the coast. On the other hand, the data showed that concentrations of accumulated metals in wedge clams from the studied localities were, as a whole, below the maximum residual levels prescribed by local and international regulations for seafood. Recent studies on commercial wedge clams from the Bulgarian Black Sea coast had already indicated that the species is generally safe for human consumption, and the new findings broadly support that assessment.

There was one notable exception. The high concentration of lead, 2.51 milligrams per kilogram, in wedge clams from Sveti Vlas in June significantly exceeded the national and European regulations for seafood, which set a limit of 1.5 milligrams per kilogram. This single exceedance at a busy resort beach underscores how localised metal contamination can be, and how a sentinel species can flag a problem that might otherwise go unnoticed in routine water or sediment sampling. Lead is among the trace metals of greatest toxicological concern, and its persistence in the marine environment makes such a signal particularly worth following up.

Taken together, the study establishes the wedge clam Donax trunculus as a suitable bioindicator for the assessment and monitoring of metal pollution in the Bulgarian Black Sea environment. As a filter feeder that plays a significant role in maintaining the functionality of marine ecosystems, the species can accumulate considerable amounts of trace metals while remaining abundant and accessible along sandy shores. For a coastline facing growing tourism and development pressure, the small clam buried in the sand may prove to be one of the most informative witnesses to the state of the Black Sea environment, turning every handful of beach sediment into a record of the metals that flow through it.

Subject of Research: Trace metal bioaccumulation in wedge clams as a bioindicator of pollution on the Bulgarian Black Sea coast

Article Title: Trace metal accumulation in tissues of wedge clams from sandy habitats of the Bulgarian Black Sea coast

Article References: Bachvarova, D., Stancheva, M., Ignatova-Ivanova, T., Ibryamova, S., Chipev, N., & Alexandrova, A. (2022). Trace metal accumulation in tissues of wedge clams from sandy habitats of the Bulgarian Black Sea coast. BioRisk, 17, 73-81. https://doi.org/10.3897/biorisk.17.77290

Image Credits: AI Generated

DOI: 10.3897/biorisk.17.77290

Keywords: wedge clam, Donax trunculus, trace metals, Black Sea, bioindicator, metal pollution, Bulgaria, ICP-OES, marine monitoring, bioaccumulation, Trace, metal

The study, published in Frontiers in Marine Science, draws on water-quality records collected between 2001 and 2024 at 13 fixed monitoring stations along Kuwait’s coastline, six inside Kuwait Bay and seven distributed along the open Gulf coast to the south. Chlorophyll-a, the pigment that gives phytoplankton their green color, is widely used as a proxy for phytoplankton biomass and as a practical indicator of the trophic status of coastal waters. Because phytoplankton respond rapidly to changes in nutrient availability, light, temperature and water-column stability, long-term chlorophyll-a records are among the most effective tools for detecting eutrophication and distinguishing gradual ecological trends from abrupt shifts in baseline conditions.

The numbers are striking. Between the pre-construction period of 2001–2012 and the construction years of 2013–2017, mean annual chlorophyll-a increased 2.9-fold at the Bay sites, rising from 1.29 to 3.7 milligrams per cubic meter, and 2.7-fold at the Gulf sites, rising from 1.16 to 3.08 milligrams per cubic meter. Annual means peaked in 2016, reaching 7.6 milligrams per cubic meter inside the Bay and 5.0 in the adjacent Gulf waters. Concentrations during 2021–2024 declined from those construction-era highs, but at 2.31 and 1.92 milligrams per cubic meter respectively they remained well above the pre-shift baseline. The system, in other words, did not simply return to normal once the construction cranes left.

Statistically, the pattern is better described as an abrupt regime shift than a steady climb. The researchers applied the non-parametric Mann–Kendall test and Sen’s slope estimator to annual mean chlorophyll-a in each region. Over the full 2001–2024 record, chlorophyll-a increased by 0.061 milligrams per cubic meter per year at the Bay sites and 0.051 at the Gulf sites, but the monotonic trend was statistically significant only inside Kuwait Bay, where the Mann–Kendall tau of 0.38 carried a p-value of 0.025. The Pettitt change-point test, a non-parametric method designed to detect a single abrupt shift in the mean of a time series, pinpointed a significant change at the 2012–2013 boundary in both regions. Crucially, when the pre- and post-change segments were analyzed separately, neither showed a significant internal trend. The positive full-series slopes therefore reflected a step-like transition to a new, higher baseline rather than continuous year-by-year accumulation.

What makes the finding genuinely puzzling is that the usual suspect, nutrient enrichment, does not line up with the timing of the shift. Multiple regression models fitted to log-transformed chlorophyll-a and seven co-measured environmental variables, including water temperature, salinity, dissolved oxygen, nitrate, nitrite, phosphate and total suspended solids, explained only 9.5 percent of chlorophyll-a variability at the Bay sites and 11.9 percent at the Gulf sites. Exploratory Pettitt tests applied to each environmental variable found no significant shift at the 2012–2013 chlorophyll-a boundary after correction for multiple testing. The abrupt ecological change, the authors conclude, cannot be explained by a corresponding increase in concurrent dissolved nutrient concentrations.

This nutrient–biomass decoupling is less paradoxical than it first appears, and it echoes patterns documented in coastal systems worldwide. Chlorophyll-a reflects accumulated phytoplankton biomass, whereas a water sample measures only the dissolved nutrient pool remaining after transport, biological uptake and internal recycling. During a shelf-sea spring bloom in the Celtic Sea, for example, rising chlorophyll-a was accompanied by rapid nitrate drawdown over days, illustrating how phytoplankton growth can deplete the very nutrients that fuel it. Monthly sampling may simply miss such short-lived dynamics. Internal nutrient recycling can also sustain production without producing elevated nitrate or phosphate readings, and longer water residence time can allow biomass to accumulate before it is exported from a coastal system. The negative association between nitrate and chlorophyll-a retained in the Gulf-site mixed-effects analysis is consistent with biological drawdown during bloom periods, and the positive dissolved-oxygen association likely reflects photosynthetic oxygen production by the phytoplankton themselves rather than oxygen driving biomass.

The temporal correspondence with the causeway is where the story becomes genuinely provocative. The Sheikh Jaber Al-Ahmad Causeway, a 36-kilometer crossing of Kuwait Bay built largely between 2013 and 2018 and opened in 2019, is among the longest marine crossings in the world. Independent field observations and hydrodynamic modelling show that the structure and associated land reclamation measurably altered the bay’s circulation. Surface-current speeds within 5 kilometers of the causeway declined by roughly 16 percent in summer and 12 percent in spring, an effect attributed to the densely arranged piers acting as a physical barrier to water exchange. Subsequent modelling found that the bay-wide average water residence time increased by 1.29 days, or 3.49 percent, while local effects ranged from a 66-day reduction to a 56-day increase depending on location. Complete renewal of Kuwait Bay, which already renews slowly, was estimated to require 150 to 320 days. Chlorophyll-a reached its highest recorded levels in 2015–2016, during the most intensive phase of construction.

The authors are careful to state that temporal correspondence does not establish causation, and they did not identify an independently documented abrupt regional hydrographic or climatic event at the 2012–2013 boundary. Longer-term hydrographic change provides relevant context but not a discrete trigger: the estimated freshwater contribution at a northern Gulf station declined from 25.6–42.5 percent in 1997 to 0.8–4.6 percent by 2012–2013, accompanied by a salinity increase of roughly six units between 2000 and 2013, a progressive decadal trend rather than a sudden event. Comparable engineering studies support the proposed mechanism, however. In Doha Bay, Qatar, land reclamation increased mean water residence time by 3–6 days between 2000 and 2020, tripling in some areas, while at Korea’s 33.9-kilometer Saemangeum seawall chlorophyll-a was consistently higher inside the enclosure and particularly elevated when sluice-gate operation restricted water exchange. Restricted exchange, these studies suggest, can promote phytoplankton biomass accumulation by extending the time nutrients and cells are retained within a semi-enclosed system.

The spatial pattern adds another layer of intrigue. The shift was detected not only inside Kuwait Bay but also at Gulf stations distributed along roughly 80 kilometers of open coastline, some within 14 to 27 kilometers of the Bay, indicating a coastal-scale response rather than a purely local one. This divergence from the wider Gulf is notable: satellite ocean-color analyses indicate that chlorophyll-a across the Gulf as a whole declined by approximately 16 percent between 2003 and 2021, with interannual variability governed largely by climate oscillations such as the Indian Ocean Dipole and the El Niño–Southern Oscillation. Kuwait’s coastal waters bucked that regional decline, pointing to local and subregional processes, including chronic nutrient inputs from urban and industrial sources and changes in circulation and retention, acting alongside broader climatic forcing. Because the Bay and adjacent Gulf waters form a hydrodynamically connected system, changes affecting circulation and biological retention need not remain confined within the Bay.

The implications reach well beyond Kuwait. The study demonstrates that routine water-quality monitoring can detect abrupt ecological change but cannot, on its own, disentangle the relative contributions of nutrient loading, biological processing and hydrodynamic modification. The authors argue that chlorophyll-a time series should be paired with phytoplankton community observations, continuous high-frequency measurements of chlorophyll-a, dissolved oxygen and temperature, event-based nutrient sampling, and direct measurements of currents and water exchange supported by hydrodynamic modelling. For future coastal engineering projects, monitoring should begin before construction and continue through and after it, with appropriate comparison sites, so that project-associated effects can be separated from broader regional variability. In one of the world’s most environmentally extreme seas, where summer temperatures reach 37 degrees Celsius and salinity exceeds 40 practical salinity units, the message is clear: when humans reshape the physics of a coastal sea, the biology may answer with an abrupt, persistent and difficult-to-reverse shift that no nutrient measurement alone could have predicted.

Subject of Research: Marine Science

Article Title: Long-term chlorophyll-a dynamics reveal abrupt baseline shifts in coastal waters of the northwestern Arabian Gulf

Article References: Abdulhussain, A. H., Alosaimi, W., & Ben-Hasan, A. (2026). Long-term chlorophyll-a dynamics reveal abrupt baseline shifts in coastal waters of the northwestern Arabian Gulf. Frontiers in Marine Science, 13, Article 1949281. https://doi.org/10.3389/fmars.2026.1949281

Image Credits: AI Generated

DOI: 10.3389/fmars.2026.1949281

Keywords: chlorophyll-a, Kuwait Bay, Arabian Gulf, phytoplankton, causeway, coastal engineering, regime shift, marine monitoring, Long-term, dynamics, reveal, abrupt

The challenge is formidable. Acetes chinensis individuals measure just three to four centimeters, and in the echograms produced by a scientific echosounder they may occupy only a handful of pixels. The team, led by Yanrui Qu and Sheng Chen of the Third Institute of Oceanography in Xiamen, found that the median shrimp bounding box in their dataset covered just 0.189 percent of the image area. Layer on high-density aggregations, overlapping echoes, reverberation, and background acoustic clutter, and conventional object-detection models struggle badly, missing weak targets or mistaking noise for shrimp.

Acoustic sensing is the right tool for the job because it does not depend on light. Coastal waters near industrial intakes are often turbid, rendering optical cameras nearly useless, while a broadband split-beam echosounder such as the SIMRAD EK80 can operate regardless of visibility. But traditional hydroacoustic analysis, which relies on echo integration and target-strength measurements, estimates aggregate biomass rather than identifying and counting individual animals. Deep-learning detection fills that gap, providing explicit spatial localization that can be combined with tracking for automated enumeration.

The researchers started with RT-DETR, a real-time detection transformer that predicts objects end-to-end without the non-maximum suppression step used by most CNN-based detectors. Its self-attention mechanism models long-range dependencies between target and background regions, which matters when shrimp echoes are embedded in heterogeneous acoustic noise. The baseline RT-DETR-L, initialized with COCO-pretrained weights, already achieved a mean average precision at an intersection-over-union threshold of 0.5, or mAP@50, of 0.818 on their echogram dataset.

Three targeted modifications pushed performance higher. First, the team added a high-resolution P2 detection layer. The original architecture predicts from feature levels P3 through P5, whose spatial resolutions drop as the network deepens, losing fine detail about tiny targets. The new P2 branch fuses upsampled P3 features with shallow backbone features, producing a 160-by-160 feature map for 640-by-640 input, four times the spatial locations of P3, so weak echo boundaries survive the downsampling process.

Second, the researchers integrated Efficient Multi-Scale Attention, or EMA, modules at three backbone stages and three neck stages. These modules adaptively reweight features along both channel and spatial dimensions, amplifying shrimp-related responses while suppressing strong but irrelevant background echoes such as scattering and sidelobe interference. Notably, ablation experiments showed that placement matters: inserting EMA before the Adaptive Intra-scale Feature Interaction module yielded the best results, with mAP@50 rising to 84.2 percent, compared with 82.9 percent when placed between the encoder’s two modules.

Third, the loss function was rebuilt. Because background pixels vastly outnumber shrimp pixels, the team adopted Focal Loss for classification, which down-weights easily classified samples so training focuses on difficult cases. For bounding-box regression they used Wise-IoU version 3, which applies a dynamic non-monotonic focusing mechanism that reduces the influence of low-quality samples on gradient updates. This matters for small boxes, where even a slight displacement of a predicted boundary causes a large drop in overlap with the ground truth. The two losses proved complementary: together they lifted mAP@50-95, a stricter localization metric, to 0.437.

The fully modified model reached a mean mAP@50 of 0.844 and mAP@50-95 of 0.437, with precision of 0.834 and recall of 0.841, all statistically significant improvements over the baseline in paired t-tests. Recall improved by nearly six percentage points, indicating the architecture is particularly good at recovering weak targets that would otherwise slip past. The gains came cheap: parameters rose only from 31.99 million to 32.21 million, and the model outperformed small-object specialists QueryDet and TridentNet while remaining competitive with heavyweight detectors like DINO and RT-DETRv2, which it beat on recall despite their higher mAP scores.

Detection alone, however, cannot count shrimp, because the same individual appears in many consecutive frames. The team therefore coupled their detector with DeepSORT, a tracking algorithm that combines a Kalman filter motion model with appearance embeddings from a Wide Residual Network, matching detections to trajectories via the Hungarian algorithm. Each confirmed trajectory counts as one shrimp. Across ten ten-minute video clips, DeepSORT achieved a counting accuracy of 83.32 percent, outperforming SORT and ByteTrack, which reached 74.88 and 81.08 percent respectively. Qualitative examples showed the improved model dropping a false-positive background response from a confidence of 0.42 to 0.18, below the detection threshold, while boosting a weak shrimp echo from 0.11 to 0.35, rescuing a missed detection.

The authors are candid about limitations. Their data came from a controlled seawater tank, and generalization to different species, depths, densities, and acoustic environments remains untested. Counting errors still arise from trajectory fragmentation, identity switches, and temporary target disappearance, and the added P2 processing slows inference relative to the baseline. Future work will incorporate temporal and multi-frequency acoustic information and more robust trajectory association. Still, the framework demonstrates that deep-learning detection can extend from fish schools to individual planktonic crustaceans, giving nuclear plant operators a practical, non-invasive way to monitor biological loading before a shrimp swarm becomes a shutdown event.

Subject of Research: Deep learning-based detection and counting of Acetes chinensis shrimp in EK80 acoustic echogram images for nuclear power plant intake monitoring

Article Title: Improved RT-DETR-based detection and counting of shrimp in EK80 acoustic echogram images

Article References: Qu, Y., Chen, S., Xue, R., & Zheng, S. (2026). Improved RT-DETR-based detection and counting of shrimp in EK80 acoustic echogram images. Frontiers in Marine Science, 13, Article 1917117. https://doi.org/10.3389/fmars.2026.1917117

Image Credits: AI Generated

DOI: 10.3389/fmars.2026.1917117

Keywords: RT-DETR, deep learning, acoustic echogram, shrimp detection, Acetes chinensis, EK80 echosounder, object detection, multi-object tracking, DeepSORT, nuclear power plant, marine monitoring, small object detection

The researchers call their model HA-DETR, short for Hierarchical Adaptive Feature Enhancement Detection Transformer. It builds on the detection transformer family of architectures, which replaced the hand-engineered region proposal pipelines of earlier detectors with an end-to-end attention mechanism that reasons about an entire image at once. Since the original DETR appeared in 2020, the family has evolved rapidly through variants such as Deformable DETR, DAB-DETR, and DINO, and recent work has shown that detection transformers can now rival or exceed the one-stage YOLO detectors that long dominated real-time applications. Liu and Jiao’s contribution is not a new backbone or a new loss function, but a set of three modules that intervene at specific points in the feature hierarchy, each tailored to the particular kind of degradation that dominates at that level.

The key observation motivating the architecture is that underwater degradations affect low-level details and high-level semantic representations differently. At the bottom of a convolutional or transformer feature pyramid, the network encodes fine-grained structure: edges, textures, and the delicate outlines of fins, tentacles, and shells. These are precisely the cues that scattering and blur destroy first. Higher up the pyramid, features become more abstract and semantic, encoding what an object is rather than where its boundaries lie, but they become vulnerable to a different problem: attention drifts toward the water column, sediment, and background clutter instead of locking onto the sparse object regions that matter. A single, uniform enhancement strategy applied to all levels therefore tends to help one level at the expense of another, which is why previous enhancement-plus-detection pipelines have delivered mixed results.

HA-DETR’s first component, the Frequency-Domain Feature Enhancement module, addresses the low-level problem by moving the computation into the frequency domain. The idea draws on a growing body of work showing that frequency analysis can separate image content in ways that spatial processing cannot: blur and haze tend to suppress high-frequency components, while object contours and textures concentrate there. By transforming features with operations related to the fast Fourier transform and adaptively reweighting frequency bands, the module restores the perception of blurred objects before the information is lost to deeper layers. This follows a broader trend in vision research, from frequency-aware transformers for image restoration to discrete cosine transform hybrids, that treats spectral information as a first-class citizen rather than a curiosity.

The second component, the Polarity-Aware Feature Interaction module, works at the opposite end of the hierarchy. Its job is to refine high-level semantic features so that the model’s attention concentrates on object regions rather than on the visually dominant but uninformative background. The term polarity refers to the module’s ability to distinguish and separately handle the opposing contributions of foreground and background signals during feature interaction, strengthening the positive evidence for objects while suppressing the negative pull of the surrounding water. In an underwater scene where a fish may occupy only a few percent of the pixels, this foreground-background asymmetry is the difference between a confident detection and a network that dutifully describes the emptiness around the fish.

The third component, the Hierarchical Attention Feature Pyramid Network, tackles the fusion problem that sits between the two extremes. Feature pyramid networks and their many descendants are the standard machinery for combining multi-scale features so that small objects benefit from high-resolution detail and large objects benefit from semantic depth. HA-DETR’s variant uses attention to optimize how information flows across scales and reinforces semantic-guided representation, ensuring that the enhanced low-level detail from the frequency module and the sharpened semantics from the polarity module are combined coherently rather than averaged into mush. The three modules together perform what the authors describe as differentiated processing and adaptive enhancement across semantic levels, allowing the network to preserve fine-grained structures and maintain robust semantic discrimination simultaneously.

The experimental evidence comes from two widely used underwater benchmarks. On DUO, a dataset of underwater object detections assembled from marine imagery with annotations for creatures such as fish, jellyfish, and sea urchins, HA-DETR achieves an average precision of 70.7 percent. On UTDAC2020, a dataset collected in a real aquaculture environment where turbidity and crowding are severe, it reaches 53.2 percent. Both figures represent the model’s superiority over competing approaches reported in the study, and the gap is particularly meaningful on the aquaculture data, where low contrast and heavy occlusion have historically dragged detection scores down. The authors report no competing interests, and the work was accepted by Cluster Computing after revisions in June 2026 and published on 18 September 2026.

What makes the approach notable beyond its benchmark numbers is the way it reframes the relationship between image enhancement and object detection. Much prior work treats underwater enhancement as a preprocessing step, using generative or physics-based models to produce a cleaner image before detection begins. That strategy can help, but it adds computational cost, can introduce artifacts that mislead the detector, and optimizes for visual quality rather than detection performance. HA-DETR instead embeds enhancement directly inside the detection architecture, applying it selectively where the feature hierarchy needs it. This echoes a recent line of research, including edge-aware DETR variants and CNN-transformer hybrids for underwater scenes, that argues detection-aware feature processing beats generic image restoration for this task.

The practical stakes are considerable. Automated monitoring of fish populations, coral reef health, and invasive species depends on reliable detection in conditions no human photographer would choose. Aquaculture operations use cameras to track feeding behavior, growth, and disease outbreaks in pens where visibility is often measured in meters. Autonomous underwater vehicles conducting infrastructure inspection or deep-sea surveying need to identify objects in real time with limited bandwidth back to shore. A detector that maintains accuracy under blur, low contrast, and occlusion extends the operational envelope of all of these systems, potentially reducing the need for human divers in hazardous or remote environments.

There are, of course, caveats. The reported results come from two datasets, and performance on other water bodies, lighting conditions, and sensor types remains to be demonstrated. Transformer-based detectors also carry computational costs that matter for the power-constrained embedded platforms on which many underwater vehicles run, although the field is moving quickly toward efficient attention designs. Still, the hierarchical adaptive principle at the heart of HA-DETR, matching the enhancement strategy to the semantic level being processed, is a general one, and the authors’ results suggest that respecting the structure of the feature hierarchy is one of the most effective ways to make machine vision work where the water refuses to cooperate.

Subject of Research: Transformer-based hierarchical adaptive feature enhancement for underwater object detection

Article Title: HA-DETR: Hierarchical adaptive feature enhancement for underwater object detection

Article References: Liu, T., & Jiao, G. (2026). HA-DETR: Hierarchical adaptive feature enhancement for underwater object detection. Cluster Computing, 29(13), Article 768. https://doi.org/10.1007/s10586-026-06577-w

Image Credits: AI Generated

DOI: 10.1007/s10586-026-06577-w

Keywords: underwater object detection, detection transformer, HA-DETR, frequency-domain enhancement, feature pyramid network, attention mechanism, computer vision, marine monitoring, aquaculture, deep learning, multi-scale feature fusion, image degradation

The scientific and legal momentum behind tools like TUNE stems largely from the European Union’s Marine Strategy Framework Directive, the landmark 2008 legislation that obliges member states to achieve Good Environmental Status in their marine waters. Under the Directive’s Descriptor 11, which covers the introduction of energy, including underwater sound, two distinct criteria apply. The first addresses low- and mid-frequency impulsive sounds, such as the hammering of pile drivers during offshore construction, while the second targets continuous low-frequency noise, dominated overwhelmingly by commercial shipping. Shipping noise is the most pervasive form of underwater noise pollution and dominates ambient sound levels below one kilohertz across much of the world’s oceans. Criterion D11C2 specifically requires monitoring programs to report sound pressure levels within decidecade frequency bands centered at 63 and 125 hertz, two narrow slices of the acoustic spectrum that serve as proxies for overall shipping-related noise exposure.

Meeting that regulatory requirement sounds straightforward, but in practice it has been anything but. Reliable estimation of these acoustic quantities depends on robust and reproducible signal processing workflows, and the software landscape has long been fragmented. Raven Pro, a mainstay of bioacoustics research, offers limited flexibility in calibration and requires a commercial license. MATLAB-based tools such as CHORUS provide advanced processing capabilities but are not distributed as freely available standalone applications. The open-source MANTA package has improved accessibility and standardization for long-term ocean soundscape analysis, but it produces outputs in hybrid millidecade bands rather than the decidecade bands adopted for Marine Strategy Framework Directive reporting. Many other packages remain proprietary or closed-source, preventing users from inspecting and verifying the algorithms that ultimately feed into official environmental assessments. TUNE was engineered to close this gap by combining calibration accuracy, usability and regulatory compliance in a single freely available package distributed under the GNU General Public License.

At its core, TUNE processes calibrated acoustic recordings in wav format and computes a suite of standard underwater noise metrics. These include broadband sound pressure level, decidecade sound pressure level, zero-to-peak sound pressure level and power spectral density estimates. The software embraces recent international standardization: it uses the term decidecade in place of one-third octave band, following the 2025 update of the relevant ISO standard, which considers the two metrics approximately equivalent, and it adopts the sound pressure symbol conventions of ISO 18405:2017 for the definitions and equations of all four metrics. Processing is carried out over user-defined time intervals called snapshots, a concept aligned with the temporal observation windows introduced in recent ISO documents. Each recording can be segmented into consecutive, skipped or overlapping snapshots, allowing fully time-resolved estimates of how noise levels rise and fall, for example as a vessel approaches and then recedes from a monitoring station.

A distinguishing technical feature of TUNE is its support for frequency-dependent calibration. Hydrophones, the underwater microphones that record ocean sound, do not respond uniformly across frequencies, and their sensitivity can vary substantially across the acoustic spectrum. Many existing tools accept only a single calibration factor, which introduces errors when the hydrophone response is uneven. TUNE instead accepts a two-column text file containing frequency values and corresponding hydrophone sensitivity, expressed either as voltage sensitivity in decibels relative to one volt per micropascal or as an equivalent calibration factor in pascals per volt, with a minimum of three frequency points required. Spectral estimates are corrected bin by bin using this sensitivity curve, improving the accuracy of the resulting sound pressure levels. The software also allows optional high-pass, low-pass or band-pass digital filtering before analysis, with filters designed to fixed passband ripple and stopband attenuation specifications, and it supports flexible FFT parameters including window type, overlap and transform length.

Internally, the software follows a modular architecture organized into five stages: user input, data import, calibration, signal processing and output generation. A single-window graphical user interface lets users configure directories, calibration files, time ranges, spectrogram settings and filtering options, and the entire parameter set can be saved and reloaded as a setup file to guarantee reproducibility across monitoring campaigns. Processing proceeds through three nested loops at the file, channel and snapshot levels, with multi-channel recordings handled channel by channel. For each snapshot, the software computes a series of Fast Fourier Transforms over signal subsegments and averages them into a single representative spectrum, then corrects that spectrum against the calibration curve. Decidecade integration is performed using the nearest available FFT bins rather than exact theoretical band edges, a pragmatic approximation that benchmarking described in the paper’s supplementary materials shows produces negligible discrepancies for standard monitoring applications, while dramatically accelerating computation. One creative addition is the option to overlay Knudsen curves for sea states zero through six, letting analysts compare measured ambient noise against classic empirical predictions of wind-driven sea-state noise.

To demonstrate the tool, the researchers analyzed a real recording of a ship passing through a harbor, captured with a Teledyne Reson TC4040 hydrophone and VP1000 preamplifier feeding a Tascam DR-100 recorder at 96 kilohertz sampling with 24-bit resolution. The ten-minute recording was processed in ten-second snapshots using 131072-point FFTs with a Hann window and fifty percent overlap, and the entire computation completed in just two minutes and thirty-five seconds. The results were vivid: mean sound pressure levels in the 63 and 125 hertz decidecade bands reached 111 and 117 decibels respectively, the mean broadband level was 123 decibels, and the maximum zero-to-peak level hit 156 decibels at 256 seconds into the recording, precisely the closest point of approach of the vessel. All four metrics rose clearly during the first half of the recording, matching the ship passage documented in the campaign logs, and every output was delivered both as figures and as numerical text files ready for statistical aggregation across recordings, sites and seasons.

The impact of the software is already visible in practice. By providing a consistent, transparent and reproducible framework for acoustic data processing, TUNE directly addresses a long-standing problem: differences in signal processing workflows can produce substantial discrepancies in reported sound pressure levels, undermining the comparability of environmental assessments across institutions and countries. Perhaps most significantly, the tool has been adopted by multiple Regional Environmental Protection Agencies in Italy, establishing a shared national framework for underwater noise monitoring and overcoming methodological fragmentation among regional programs. This software-driven harmonization improves the consistency of reported indicators, increases transparency in reporting and supports the coordinated implementation of national monitoring obligations under the Marine Strategy Framework Directive. Because TUNE can process long-term acoustic datasets and generate decidecade-band time series, it also supports the estimation of baseline conditions, trend analysis and threshold exceedance assessments in line with recent European Commission guidance on threshold values.

The developers see TUNE as a foundation rather than a finished product. Future versions will extend the range of computed metrics to include the Sound Exposure Level, which would allow the software to cover the Directive’s impulsive-sound criterion in addition to continuous noise, along with greater flexibility in temporal analysis. Beyond European regulatory monitoring, the authors note that the tool’s applicability extends to a broad range of marine and freshwater acoustic studies, from low-noise calibration sites to research linking low-frequency noise shifts to changes in fin whale song. In a field where proprietary licenses and closed algorithms have often stood between environmental agencies and their legal obligations, TUNE represents a quietly radical proposition: that the measurement of a pollutant threatening ocean life should be accurate, transparent and free for anyone who needs it.

Subject of Research: A free standalone software tool for standardized evaluation of underwater noise pollution under the Marine Strategy Framework Directive

Article Title: TUNE: a Tool for Underwater Noise Evaluation

Article References: Caradonna, V., Buogo, S., Azzellino, A., Vicinanza, D., & Borsani, J. F. (2026). TUNE: a Tool for Underwater Noise Evaluation. SoftwareX, 35, Article 103008. https://doi.org/10.1016/j.softx.2026.103008

Image Credits: AI Generated

DOI: 10.1016/j.softx.2026.103008

Keywords: underwater noise, TUNE software, Marine Strategy Framework Directive, shipping noise, hydrophone calibration, sound pressure level, decidecade bands, power spectral density, ocean acoustics, marine monitoring, MATLAB, environmental protection