The new study, published in the journal Coral Reefs, was conducted by a team from the Australian Institute of Marine Science working across six macroalgal-influenced reef sites around the Keppel Islands, on Woppaburra sea Country in Queensland, Australia. The researchers tested whether a commercial, biocide-free foul-release coating, or FRC, applied to ceramic coral seeding devices could reduce fouling accumulation and improve the chances of survival for juvenile corals of the branching species Acropora millepora. The results were striking: devices treated with the coating carried roughly 24 percent fouling after 48 weeks in the water, compared with about 67 percent on untreated controls, and spat survival rose from roughly 22 percent to about 49 percent.

Foul-release coatings work very differently from the biocidal antifouling paints used on ships’ hulls. Rather than leaching toxic copper or organotin compounds, which are known to harm coral fertilisation and larval metamorphosis, these coatings rely on hydrophobic or amphiphilic surface chemistry that simply weakens the adhesion strength of fouling organisms. Algae and invertebrates struggle to grip the slick surface, and much of what does attach is easily sloughed off by water movement. That property has made FRCs increasingly popular in aquaculture and maritime applications, and earlier work by the same group had shown they reduced fouling on ceramic seeding devices by up to 70 percent at healthier, coral-dominated mid-shelf reefs, with a modest boost to spat survival of up to 12 percent.

What remained uncertain was whether such benefits would hold up under the far harsher conditions of degraded inshore reefs, where high macroalgal biomass, eutrophication and sedimentation create a fusillade of biological pressure on any artificial surface. To find out, the team collected gametes from 37 gravid Acropora millepora colonies around the Keppel Islands during the October 2022 spawning event, fertilised them on board an aquaculture-configured vessel, and reared the larvae in culture tanks. Once more than 70 percent of the larvae had reached settlement competency, roughly five days after fertilisation, they were induced to settle onto preconditioned concrete tabs carrying a crustose coralline algae biofilm, at a density of around 24,000 larvae per tank.

The settled spat, cut onto small 14 by 14 millimetre settlement tabs, were then slotted into alumina seeding devices, 216 in total, half coated with the commercial FRC Hempasil 77300 and half left as controls. Divers deployed the devices at six sites across the Keppels in December 2022 and retrieved them 48 weeks later. The sites spanned a gradient of benthic regimes: three were dominated by canopy-forming Sargassum macroalgae covering nearly two-thirds of the substrate, while the others were characterised by encrusting Lobophora and coral assemblages dominated by Montipora. At retrieval, divers recorded each device’s burial status, photographed the surrounding benthos, and brought the fouled devices back to the laboratory for detailed image analysis of fouling cover and spat survival.

The scale of the fouling difference was dramatic. The largest treatment effect appeared at Humpy, where control devices averaged nearly 72 percent fouling cover while coated devices carried only about 10 percent. Across all sites, the dominant foulers were crustose coralline algae and brown and red algae, both of which were substantially rarer on coated devices. Crucially, survival declined steeply with increasing fouling on the settlement tabs themselves: the model predicted survival of around 43 percent on completely clean tabs but under 4 percent where fouling approached total cover. Burial compounded the problem, with survival dropping to less than 1 percent on fully buried devices, whether they were buried by sediment, rubble, macroalgae or overgrowing coral.

Interestingly, the surrounding benthic community shaped outcomes in ways that were not always intuitive. Survival was generally higher at Sargassum-dominated sites than where understorey macroalgae such as Lobophora and Caulerpa, or encrusting corals, prevailed, possibly because these lower-growing competitors and sediment-retentive assemblages creep directly onto devices. Sediment accumulation of up to 25 percent on tabs at one site corresponded with sharply reduced survival regardless of treatment. Yet even across these environmental gradients, the FRC benefit remained consistent: coated devices maintained predicted survival above 30 percent across most benthic assemblages, while control survival fell to as little as 2 percent at coral-dominated locations. Posterior estimates suggested a 38 percent higher probability of finding live spat on FRC devices overall.

There was also a hint that the coating helped in an unexpected way, by discouraging burial. Fully buried devices were consistently less common among FRC treatments than controls, and the authors suggest that by limiting initial fouling accumulation, the slick surfaces may deny secondary colonisers, encrusting algae and corals the foothold they need to slowly overgrow and entomb the devices. Spat on coated devices also grew larger, with tissue covering up to about 24 percent of tabs compared with roughly 7 percent on controls, likely because reduced competitive pressure freed resources for growth, accelerating the recruits towards the size refuge at which they become far less vulnerable to overgrowth.

The findings carry practical weight for a restoration field that is racing against repeated mass bleaching. The 2024 bleaching event alone cut coral cover by more than 20 percent across the Great Barrier Reef and caused mortality approaching 90 percent on some reefs around Lizard Island, while inshore systems such as the Keppels face added pressure from flood plumes, turbidity and strong tidal extremes. Paradoxically, the same turbid conditions that buffer adult colonies against heat stress may worsen sediment deposition and fouling pressure at the millimetre scale where spat live. The Keppel Islands’ Acropora populations have repeatedly bounced back from six major bleaching events in three decades, but the authors caution that resilience at the colony level does not automatically protect the fragile post-settlement stages that determine whether a population can rebuild.

The team is quick to note that foul-release coatings are not a panacea. They mitigate one critical bottleneck, competition and overgrowth at the settlement surface, but they cannot override heavy sedimentation or poor site selection, and survival still fell sharply where tabs accumulated silt or were buried. The authors recommend pairing FRC-treated devices with site-prioritisation frameworks that minimise sediment and burial risk, testing coatings applied even closer to the settlement surface, trialling biologically inert settlement substrates, and refining device geometry to improve hydrodynamic performance. Longer-term monitoring will also be needed to capture seasonal swings in macroalgal cover. Still, as a scalable, non-toxic intervention, the approach offers something restoration has badly needed: a way to keep young corals alive long enough on degraded inshore reefs to give them a fighting chance of reaching adulthood.

Subject of Research: Testing biocide-free foul-release coatings on coral seeding devices to improve spat survival on macroalgal-dominated inshore reefs

Article Title: Coral spat survival on inshore reefs is enhanced by foul-release-coated seeding devices

Article References: Coral spat survival on inshore reefs is enhanced by foul-release-coated seeding devices. (n.d.). https://doi.org/10.1007/s00338-026-02953-5

Image Credits: AI Generated

DOI: 10.1007/s00338-026-02953-5

Keywords: coral restoration, coral spat survival, foul-release coating, Great Barrier Reef, Keppel Islands, larval seeding, macroalgae, biofouling, Acropora millepora, sedimentation, reef rehabilitation, early life-stage bottleneck

The research, conducted by J. S. Shyam Mohan of Lincoln University College in Malaysia and Ankur Dumka of the Women Institute of Technology in Dehradun, India, addresses a problem that has frustrated automated marine inspection for years. While remotely operated vehicles and fixed cameras can gather enormous quantities of visual data from offshore platforms, ship hulls and submerged infrastructure, turning those images into reliable maps of fouling coverage has remained stubbornly difficult. Underwater imagery suffers from light attenuation, scattering, non-uniform illumination and color cast, all of which degrade the very details a computer vision system needs. On top of that, annotated underwater datasets are scarce, and the fouling classes they contain are often heavily imbalanced—common organisms dominate while rare but structurally significant species barely register.

Existing approaches to automated biofouling analysis have mostly settled for coarse image-level classification, telling an operator whether a scene contains fouling but not where it lies or how extensive it is. Pixel-level segmentation is far more demanding, requiring the model to assign a label to every individual pixel, and previous attempts have struggled with both accuracy and stability when confronted with the high-dimensional, degraded data characteristic of real underwater surveys. The authors of the new study argue that this gap is precisely where their method makes its contribution, offering a model that is accurate, robust and interpretable at the same time rather than trading one quality for another.

The researchers call their approach MRPixelDNet, a unified framework that combines four mutually reinforcing components: underwater image preprocessing, synthetic data augmentation, a deep network architecture built around mutual reinforcement between pixels, and an explainable visual interpretation layer. Rather than treating these stages as separate pipeline steps, the design is intended to let each component strengthen the others—cleaner images feed better segmentation, better segmentation produces more trustworthy interpretations, and those interpretations guide further refinement of the model’s attention across pixels. The novelty, the authors contend, lies precisely in this unified design, which simultaneously improves segmentation performance, resilience to underwater visual distortion and the explainability of predictions.

The preprocessing stage tackles the optical pathology of the underwater environment directly, enhancing texture and color information that would otherwise be lost to the water column. This matters because fouling organisms differ in fine surface texture and hue; if those cues are blurred or shifted, a segmentation network has little to work with. The augmentation stage addresses the data scarcity problem by expanding the training set with synthetic variations, allowing the network to learn from a broader range of degradation conditions, organism appearances and scene compositions than real annotated data alone could ever provide. Together, these measures give the deep network a far cleaner and more diverse foundation than raw underwater footage would allow.

At the heart of the system is the pixel-level deep network itself, which the authors describe as a mutually reinforcing architecture informed by non-local neural network principles. Instead of classifying each pixel in isolation, the network considers relationships across the entire image, allowing distant but contextually related regions to inform one another’s predictions. A patch of algae on one side of a frame can help the model correctly identify a similar-looking patch on the other side, and the boundaries between fouling species and bare substrate become sharper because the network weighs global structure alongside local appearance. The architecture also generates pixel-by-pixel masks, producing detailed maps that show exactly which regions of a submerged surface are colonized and which remain clear.

What sets the work apart from many deep learning pipelines, the authors emphasize, is its commitment to explainability. For infrastructure managers who must act on inspection results—scheduling cleaning operations, prioritizing repairs, certifying structures for continued service—a black-box prediction is of limited use. The explainable component of MRPixelDNet produces visual interpretations of its own predictions, highlighting the evidence the model relied upon when it marked a region as fouled. This transparency makes it easier for human inspectors to verify results, catch errors and build justified confidence in automated reporting, a requirement that becomes especially important when decisions carry safety and financial consequences for expensive offshore assets.

The experimental validation was carried out on underwater biofouling datasets, where the proposed method was benchmarked against state-of-the-art segmentation techniques. The results were consistent and substantial. Intersection over Union, a strict measure of how well the predicted segmentation overlaps the ground truth, improved by approximately six to ten percent, while the Dice coefficient, another standard overlap metric, rose by roughly five to eight percent. Precision, recall and accuracy also improved consistently, suggesting the gains were not confined to a single metric or a favorable subset of images but reflected a genuinely stronger ability to locate and delineate fouling under real imaging conditions.

The implications extend well beyond the metrics themselves. Accurate, interpretable maps of biofouling coverage could transform structural health monitoring in marine environments, turning what is today a labor-intensive and subjective inspection process into a scalable, data-driven workflow. Fouling is not merely a cosmetic nuisance; by altering surface roughness and drag it increases fuel consumption for vessels, and by trapping moisture and corrosive agents against structures it accelerates the degradation of assets worth billions of dollars. A system that can quantify fouling reliably, at the pixel level, in degraded imagery, and justify its own judgments, offers operators a way to detect problems earlier and allocate maintenance resources more rationally across fleets, platforms and coastal installations.

The authors present the work as evidence that segmentation accuracy, robustness and explainability need not compete—under the right architecture, each can reinforce the others in service of a practical monitoring goal. Published open access, with the accepted manuscript already citable ahead of its final version of record, the study offers marine engineers and computer vision researchers alike a detailed blueprint for bringing trustworthy automated inspection beneath the waves, where the most persistent threats to infrastructure grow silently, one settling organism at a time.

Subject of Research: Explainable pixel-level deep learning for marine biofouling segmentation and structural health monitoring

Article Title: A mutually reinforcing pixel-level deep network-driven explainable AI for biofouling segmentation and structural health monitoring in marine environments

Article References: Shyam Mohan, J. S., & Dumka, A. (2026). A mutually reinforcing pixel-level deep network-driven explainable AI for biofouling segmentation and structural health monitoring in marine environments. Journal of Big Data. https://doi.org/10.1186/s40537-026-01568-5

Image Credits: AI Generated

DOI: 10.1186/s40537-026-01568-5

Keywords: biofouling, marine engineering, explainable AI, deep learning, image segmentation, underwater imaging, structural health monitoring, computer vision, non-local neural networks, data augmentation, mutually, reinforcing

Researchers led by Nadarajan Viju of AMET University in Chennai, together with colleagues at Smykon Biotech and King Abdulaziz University, took a genomics-first approach to antifouling discovery. Rather than screening crude bacterial extracts, they mined the whole genomes of 30 marine bacteria belonging to the genera Pseudovibrio, Pseudomonas and Bacillus, all retrieved from the NCBI GenBank database. Using the open-source gene mining platform AntiSMASH, they mapped the biosynthetic gene clusters, or BGCs, the sets of genes that encode the enzymatic machinery for producing secondary metabolites. Each genome carried between 5 and 17 such clusters, yielding 278 clusters in total across the 30 organisms analysed.

The clusters fell into familiar categories: non-ribosomal peptide synthetases accounted for 28 percent, ribosomally synthesized and post-translationally modified peptides for 15 percent, terpenes for 11 percent, polyketide synthases for 9 percent, NRPS-PKS hybrids for 5 percent, siderophores for 6 percent, and a mixed bag of other clusters for the remaining 26 percent. Every species produced its own characteristic repertoire, with Pseudovibrio strains predicted to make compounds such as pyoverdin, rimosamide, prodigiosin and pseudaminic acid, Pseudomonas strains predicted to yield viscosin, pyoluteorin, mitomycin and ectoine among others, and Bacillus strains carrying genes for surfactin, bacillaene, difficidin, bacilysin and lichenysin, to name only a few.

But one compound stood out for its ubiquity. Fengycin, a cyclic lipopeptide encoded by a 22,502-base-pair non-ribosomal peptide synthetase cluster, was predicted in 26 of the 30 genomes, or 86.66 percent of all strains analysed. It was present in every one of the ten Pseudovibrio genomes, in 90 percent of the Pseudomonas genomes and in 70 percent of the Bacillus genomes. The authors argue that this broad distribution suggests fengycin acts as a conserved ecological trait, a chemical weapon that helps its producers compete for space and nutrients, colonize surfaces and survive in densely populated marine biofilms. In other words, the compound may represent a chemical defence strategy honed by evolution in exactly the kind of surface-bound microbial communities that kick off biofouling.

To test whether fengycin could plausibly block fouling at the molecular level, the team turned to structure-based virtual screening. They docked fengycin, obtained from the PubChem database, against six target proteins retrieved from the Protein Data Bank: the penicillin-binding protein of Acinetobacter baumannii and the lanosterol 14-alpha demethylase of Candida albicans as antimicrobial targets; a bacterial cell surface protein and the acyl-homoserine lactone synthase LasI, a quorum-sensing enzyme, as antibiofilm targets; and the barnacle cement protein and the mussel proximal thread matrix protein, which mediate larval adhesion, as antifouling targets. The docking was performed in PyRx using AutoDock Vina with an exhaustiveness value of eight, generating nine binding poses per ligand, with penicillin G, fluconazole and the commercial biocide DCOIT serving as reference ligands.

The results were striking. Fengycin achieved predicted binding affinities well ahead of every reference compound at every target. Against the penicillin-binding protein it scored minus 11.4 kilocalories per mole, compared with minus 6.7 for DCOIT. Against the fungal lanosterol demethylase it reached minus 13.3 kilocalories per mole, far beyond fluconazole’s minus 5.2. Its scores against the bacterial cell surface protein, the AHL synthase LasI, the barnacle cement protein and the mussel byssus protein were minus 15.3, minus 14.0, minus 14.2 and minus 13.1 kilocalories per mole respectively, while DCOIT managed only minus 7.3, minus 6.1, minus 5.5 and minus 5.61. Triplicate docking runs showed standard deviations below 0.5 kilocalories per mole, and redocking validation returned root-mean-square deviation values below the accepted threshold of 2.0 angstroms, confirming that the protocol reliably reproduced known binding orientations.

Visualization of the ligand-receptor complexes in PyMOL and BIOVIA Discovery Studio Visualizer explained the affinity. Fengycin formed extensive networks of polar and non-polar contacts: hydrogen bonds with residues such as GLU67 and LYS137 in the penicillin-binding protein, ARG381 in the fungal demethylase, SER374, SER520 and THR662 in the cell surface protein, HIS399 in LasI, GLN165 and GLU192 in the barnacle cement protein, and multiple residues in the mussel thread matrix protein, alongside dense hydrophobic contacts throughout each pocket. The researchers attribute this versatility to fengycin’s amphiphilic architecture, a rigid cyclic peptide ring fused to a hydrophobic beta-hydroxy fatty acid chain. The ring supplies multiple hydrogen bond donors and acceptors while limiting entropic penalties on binding, and the lipid tail drives van der Waals interactions within non-polar regions, allowing the molecule to engage diverse targets simultaneously.

The computational findings dovetail with decades of experimental literature on fengycin’s bioactivity. Studies have documented its antifungal action against Fusarium moniliforme, Botrytis cinerea and Magnaporthe grisea, its antibacterial effects against Xanthomonas and Pseudomonas pathogens, and, notably, its ability to disrupt quorum sensing in Staphylococcus aureus, a result reported in Nature in 2018 that aligns closely with the strong predicted interaction between fengycin and the quorum-sensing enzyme LasI in the present study. Because biofilms serve as settlement cues for many macrofoulers, a compound that interferes with both microbial adhesion and quorum sensing could in principle suppress fouling at multiple stages, from initial colonization through larval recruitment, in contrast to traditional biocides that simply poison organisms indiscriminately.

The authors are careful to frame the work as hypothesis-generating rather than conclusive. Molecular docking offers a static, simplified picture of binding that ignores protein flexibility, solvent effects, bioavailability, toxicity and cellular context, and docking scores are sensitive to ligand size, so fengycin’s large surface area may inflate its apparent advantage over small molecules like DCOIT. BGC predictions similarly depend on genome assembly quality and database annotations, and the predicted metabolites remain putative until chemically verified. The comparison with DCOIT, a biocide with documented environmental concerns of its own, must likewise be treated with caution given the compounds’ very different molecular dimensions and physicochemical properties.

Even so, the study sketches a compelling vision for the future of antifouling technology. Fengycin is biodegradable, reportedly low in toxicity, stable across ranges of temperature, pH and salinity, and potentially effective at low concentrations, making it an attractive starting point for eco-friendly coatings. More broadly, the work demonstrates that coupling genome mining with molecular docking can rapidly link biosynthetic potential to plausible biological function, providing a genomics-guided framework for prioritizing marine natural products before any laboratory assay is run. The next step is clear: in vitro and in vivo validation against real fouling organisms, together with toxicity testing, will determine whether this bacterial chemical weapon can be translated into the antifouling paints of a post-TBT world.

The study’s genome-guided strategy reflects a broader shift in natural products research. Traditional antifouling discovery relied on collecting marine organisms, extracting compounds and testing them laboriously in assays, a process that is slow, expensive and often non-specific. By contrast, mining publicly available genomes with tools like AntiSMASH allows researchers to survey the biosynthetic potential of dozens of organisms computationally before committing laboratory resources, prioritizing the most promising candidates for synthesis and testing.

Fengycin itself is well characterized biochemically. It belongs to a family of lipopeptides produced by Bacillus species alongside surfactin and iturin, and its cyclic peptide ring is assembled by large non-ribosomal peptide synthetase enzymes rather than by ribosomes, which permits the incorporation of unusual amino acids and contributes to its structural diversity. Its amphiphilic nature, combining a polar peptide head with a fatty acid tail, underlies both its surface activity and its ability to interact with biological membranes, properties that have made lipopeptides of interest as biocontrol agents in agriculture as well as in marine applications.

The choice of target proteins in the docking analysis also illustrates how antifouling can be attacked at distinct biological stages. Blocking quorum-sensing enzymes such as LasI could prevent bacteria from coordinating biofilm formation, while disrupting adhesion proteins used by barnacle larvae and mussels could stop macrofoulers from settling on a surface already colonized by microbes. A single compound active against both microbial and invertebrate targets would therefore offer multi-stage protection, a property conventional biocides achieve only through broad toxicity. The authors emphasize, however, that docking predictions must now be followed by laboratory and field validation before any practical coating can emerge.

Subject of Research: In silico prediction of the antifouling potential of the marine bacterial cyclic lipopeptide fengycin

Article Title: An in silico antifouling potential of fengycin, a cyclic lipopeptide produced by marine bacteria

Article References: Viju, N., Vijayaraghavan, P., Satheesh, S., & Jayaprakashvel, M. (2026). An in silico antifouling potential of fengycin, a cyclic lipopeptide produced by marine bacteria. Discover Oceans, 3(1), Article 51. https://doi.org/10.1007/s44289-026-00164-y

Image Credits: AI Generated

DOI: 10.1007/s44289-026-00164-y

Keywords: marine bacteria, fengycin, cyclic lipopeptide, biosynthetic gene clusters, antifouling, antibiofilm, molecular docking, AntiSMASH, biofouling, quorum sensing, natural products, silico