The findings, published in the Journal of Phycology, come at a pivotal moment. Sugar kelp is one of the dominant species in Maine’s kelp forests, providing food, habitat, and clean water that support a rich marine ecosystem. It is also the biological bedrock of the state’s burgeoning kelp aquaculture industry, the largest of its kind in the United States, which relies on the annual harvest of reproductive tissue from wild kelp populations. As ocean warming continues to drive the steady decline of these forests, understanding the genetic architecture of what remains has become an urgent scientific and economic priority.

Senior Research Scientist Doug Rasher, the study’s senior author, and his team have published several previous studies documenting how rising water temperatures are eroding Maine’s kelp forests. That loss has cascading impacts on coastal ecosystems and threatens the wild resource upon which the aquaculture industry depends. Responding effectively, the researchers argue, requires a detailed map of the structure and potential genetic barriers among the remaining forests.

“We wanted to explore questions of genetic diversity within the state of Maine because that’s the spatial scale that’s relevant to managing wild kelp forests, developing effective restoration programs, and establishing best practices in aquaculture,” Rasher said. “We brought it down to the scale that really matters from a management perspective.”

To build that map, the team collected genetic material from sugar kelp at 11 sites spanning Maine’s “outer coast.” These locations were chosen deliberately: they are far enough offshore to be subject to the oceanographic forces that influence the entire coastline, yet close enough to shore to be plausible future sources of reproductive tissue for the aquaculture industry. The analysis revealed moderate levels of genetic diversity within each study site and identified at least four distinct populations distributed along the coast.

The evidence suggests that genetic mixing occurs readily between sites within each population, presumably driven by currents dispersing kelp spores during their early life stage. But kelp are mobile for only a very brief window of time before settling on the seafloor for the remainder of their life cycle. Even with the strong currents that characterize the Gulf of Maine, spores seemingly do not travel far, and there appears to be little gene transfer between the four populations.

“Having distinct populations means there are different genetic signatures along our coast, and some populations could have individuals that are more or less suited to thrive in varied environments,” said lead author Rene Francolini, a former University of Maine PhD student in Rasher’s lab. “That’s important for restoration work or when we think about farmers who might collect reproductive tissue in one location and outplant seed in another.”

That distinction carries real-world consequences. Previous research has shown that genetic diversity can bolster resilience to marine heat waves, providing populations with the raw material to adapt as conditions shift. It also supplies critical variation for selective breeding in aquaculture, where traits such as growth rate, thermal tolerance, and morphology can determine the success of a farm. If farmers harvest reproductive tissue from one genetically distinct population and deploy the resulting seed string in waters dominated by another, they could inadvertently disrupt local adaptation—or, conversely, miss an opportunity to match the best-suited genotypes to the right environments.

Earlier studies had examined kelp population genetics at both a much larger scale, across all of New England, and at a very fine scale between adjacent bays. This study is the first to treat the Maine coast as a holistic unit, filling a critical gap between those extremes. “There’s clearly a wealth of genetic variation along the coast that needs to be considered in management and restoration and that holds great potential for innovation in aquaculture,” Rasher said.

A technological breakthrough made the study possible. Last year, another research team published the first complete sugar kelp genome—a reference resource that enabled Rasher’s group to map genetically distinct populations with far greater precision than previous marker-based approaches. The genome also opens the door to identifying specific genes unique to each population, genes that might help wild kelp forests adapt to warming waters or prove valuable in an industry setting.

“With a complete genome available, we can actually identify genes of interest and examine whether any of these genes are uniquely expressed in a given population,” Francolini said. “We have yet to be able to connect these genes to a specific function, but when we do, we will be able to pinpoint kelp variants that are uniquely strong candidates for improving kelp conservation, restoration, and aquaculture outcomes.”

The implications extend well beyond Maine. Kelp forests worldwide are retreating in the face of marine heat waves, and restoration practitioners are increasingly turning to genetics to guide their efforts—selecting donor populations that harbor heat-tolerant variants, avoiding outplanting genotypes poorly matched to local conditions, and preserving the evolutionary potential of fragmented populations. Maine’s sugar kelp, with its newly resolved population structure and a complete genome to work from, offers a template for how such efforts might proceed elsewhere.

For the aquaculture industry, the study arrives as the sector matures. Maine’s kelp farmers have built a regenerative ocean farming economy on the annual wild harvest of sorus tissue—the reproductive material from which farm seed is propagated. The new findings suggest that sourcing decisions, long made largely on geography and convenience, could be refined with genetic information: matching seed sources to outplanting sites within the same population, or deliberately drawing on genotypes from populations whose unique gene variants confer advantages in particular environments.

The research was supported by the NSF Established Program to Stimulate Competitive Research (Grant #OIA-1849227), the Louise H. & David S. Ingalls Foundation, Maine Sea Grant, and the Nature Conservancy. Co-authors include Research Scientist Robin Sleith of Bigelow Laboratory, as well as Kristina Cammen and Damian Brady of the University of Maine.

As warming continues to reshape the Gulf of Maine, the genetic diversity cataloged in this study represents both a warning and an opportunity—a reservoir of variation that, if managed wisely at the regional scale the researchers advocate, could help kelp forests endure and even thrive in the decades ahead.

The research, led by Professor Shinichi Miyagawa of the Department of Biological Science and Technology at Tokyo University of Science and co-authored by Professor Masaru Ihara of Kochi University, tackles a long-standing gap in aquatic toxicology. Scientists have known for years that fish exposed to wastewater-contaminated water can exhibit altered behavior—changes in schooling, aggression, feeding, and reproduction. But observing a behavioral change tells you little about the mechanism producing it. Antidepressants primarily work by targeting proteins known as monoamine transporters, molecular pumps embedded in the membranes of neurons that sweep serotonin, dopamine, and norepinephrine back out of the synapse after a signal is fired. The three principal members of this family are the serotonin transporter (SERT), the dopamine transporter (DAT), and the norepinephrine transporter (NET). Block these transporters, and the chemical signals linger longer, amplifying mood-related signaling—which is exactly the therapeutic goal in humans, and potentially a profound disruption in fish.

Fish possess their own versions of these transporters, but whether those fish proteins respond to human-targeted pharmaceuticals in the same way as the human versions has remained largely unexplored. Compounding the uncertainty, most mechanistic work to date has focused on a narrow set of model species, leaving open the question of whether pharmaceutical sensitivity is a general feature of fish biology or a quirk of a few laboratory favorites. To answer this, the Japanese team cast a wider evolutionary net, selecting two species separated by substantial evolutionary distance: the medaka (Oryzias latipes), a small rice fish that is a staple of Asian molecular biology laboratories, and the ayu (Plecoglossus altivelis), a commercially and ecologically important migratory fish found in Japanese rivers and coastal waters. If two such distantly related species show similar drug sensitivities, the argument goes, the pattern is likely to hold across much of the fish tree of life.

The experimental approach was elegantly mechanistic. Rather than dosing whole animals and watching what happened, the researchers first identified and cloned the genes encoding DAT, NET, and—notably—two distinct forms of the serotonin transporter, labeled SERTa and SERTb, from both species. This genetic duplication of the serotonin transporter in fish, absent from mammals, immediately raised the question of which version, if either, responds to antidepressants. The team then expressed these cloned transporter genes in cultured human cells, producing the fish proteins in a controlled laboratory setting. By flooding these cells with a fluorescent marker that the transporters normally scoop up, and then adding antidepressants to the mix, the researchers could measure precisely how effectively each drug jammed each transporter: the more a drug inhibited uptake of the fluorescent probe, the more potent its interaction with that particular protein.

The first major revelation concerned the two serotonin transporter types. Across both species, SERTa proved consistently and dramatically more sensitive to antidepressants than SERTb. This asymmetry made evolutionary sense when the researchers examined the sequences. Human SERT belongs squarely to the SERTa lineage, while SERTb—the duplicated fish-specific copy—carries substitutions at several amino acid positions known to be critical for antidepressant binding. In other words, the duplicate copy that fish uniquely possess appears to have drifted away from the drug-binding architecture that pharmaceutical designers exploited when creating human medications. The fish transporter most similar to our own, SERTa, is the one that remains exquisitely drug-sensitive.

The second revelation was more surprising, and more troubling. When the team compared fish SERTa directly with human SERT, the fish transporter frequently responded to lower drug concentrations—sometimes requiring more than ten times less drug to achieve the same degree of inhibition. This means that at any given environmental concentration of an antidepressant, the molecular machinery governing serotonin signaling in a fish brain is likely to be affected more strongly than the corresponding machinery in a human. And the surprises did not stop there. Several drugs not classically regarded as acting on these transporter proteins in humans still bound to and inhibited the fish versions, hinting at pharmacological side effects in aquatic wildlife that could never have been predicted from human pharmacology textbooks alone.

Perhaps the most consequential finding is the one that bridges the laboratory and the real world. The concentrations of antidepressants needed to block medaka SERTa in these cell-based assays overlapped directly with concentrations already measured in polluted waterways. Duloxetine, fluoxetine, citalopram, and paroxetine—all heavily prescribed medications—inhibited the medaka serotonin transporter at concentrations ranging from a few hundred nanograms per liter up to roughly 1,300 nanograms per liter, figures that sit squarely within the range documented in contaminated rivers and effluent-affected waters around the world. This is not a case of a laboratory effect requiring doses orders of magnitude above environmental relevance. The molecular target and the environmental exposure exist in the same quantitative universe.

“By demonstrating that key molecular targets in fish can be more sensitive than their human equivalents, our work offers crucial insights into the potential risks of pharmaceutical exposure to aquatic wildlife,” Professor Miyagawa noted. The implications ripple outward from molecular pharmacology into ecology. Serotonin signaling in fish influences a wide array of behaviors, including foraging, predator avoidance, shoaling, and reproductive courtship. Chronic inhibition of serotonin reuptake at environmentally realistic concentrations could, in principle, subtly rewire these behaviors across entire populations, with downstream consequences for survival and reproduction that laboratory assays of transporter inhibition cannot yet capture.

This is precisely where the study’s choice of species strengthens its message. Because medaka and ayu occupy distant branches of the fish evolutionary tree, yet display a shared pattern of heightened transporter sensitivity, the researchers argue that this vulnerability is unlikely to be a species-specific anomaly. The pharmacological architecture of the fish monoamine system—the drug-sensitive SERTa lineage, in particular—appears to be a shared ancestral trait, meaning that salmon, carp, trout, and countless other species may carry similarly sensitive molecular targets swimming through their synapses.

The research also carries a direct message for regulators and water quality authorities. Environmental monitoring programs currently face an impossible task: tracking thousands of pharmaceuticals with limited resources. Molecular sensitivity data of the kind generated by Miyagawa and Ihara’s team offers a rational basis for triage, allowing regulators to prioritize the compounds most likely to cause harm at environmentally relevant concentrations. “Our research provides a vital scientific basis for prioritizing specific pharmaceuticals in environmental monitoring programs and for deriving more protective, species-specific risk thresholds in water quality guidelines,” Miyagawa explained. In other words, safety thresholds derived from human pharmacology—or even from a single fish model—may systematically underestimate risk for the broader aquatic community.

Important questions remain. The present work was conducted in engineered human cells expressing fish transporters, a powerful system for isolating molecular interactions but one that strips away the complexity of a living animal—metabolism, blood-brain barriers, mixtures of co-occurring drugs, and compensatory physiological responses. The authors themselves emphasize that future in vivo studies examining environmentally realistic exposure scenarios and realistic pharmaceutical cocktails will be essential to determine how these molecular effects translate into measurable biological and ecological outcomes. Still, the study delivers a clear and sobering headline: the molecular locks that antidepressants were built to pick exist in fish brains too, and in fish, at least some of those locks turn more easily. As global antidepressant use continues to climb, the invisible pharmacological fingerprint of human medicine is being written into the neurochemistry of the animals that share our water—and their brains, this research suggests, may be reading that fingerprint more clearly than we ever imagined.

Now, that radical makeover is rewriting what scientists believe is possible during the course of ordinary animal development. A team of researchers led by Stanford University has found strong evidence that the majority of cells in this worm’s larval body are not discarded and rebuilt, nor do they simply carry over their old jobs into adulthood. Instead, most of them are actively reprogrammed—stripped of their larval identities and retooled to perform entirely new functions in the adult organism. The finding, published in Nature Communications, marks the first time extensive cellular reprogramming has been documented during normal development in an animal with a bilateral body plan, the body architecture shared by humans and most familiar animals.

“Reprogramming is a bit of an exotic fruit in developmental biology,” said Christopher Lowe, senior author of the study and a professor of biology in the Stanford School of Humanities and Sciences. “Generally, we used to think that as cells develop, they become increasingly restricted in their function. But when we start looking at more animals that go through metamorphosis, they may reveal that reprogramming is a much more common feature of development.”

For decades, biologists have been divided about what actually happens inside a larva as it metamorphoses. Roughly 80 percent of animal species undergo some form of metamorphosis—the stepped development that carries an animal from egg to larva to adult, a process familiar to anyone who has watched a caterpillar become a butterfly or a tadpole become a frog. Yet despite how common the phenomenon is, the cellular mechanics underlying it have remained stubbornly opaque. Two competing ideas dominated the literature. One held that larval cells largely die off during metamorphosis, and the adult body is constructed from newly generated cells. The other suggested continuity of function: a larval skin cell would become an adult skin cell, a larval neuron would remain an adult neuron, with each lineage simply growing and refining its original role.

The Stanford-led study suggests both models are, at least for this worm, largely wrong.

Paul Bump, then a doctoral student in Lowe’s lab at Stanford’s Hopkins Marine Station and now an assistant professor at Pomona College, led the effort to trace the fate of cells through the worm’s transformation. The team performed single-cell RNA sequencing on more than 87,000 individual cells, sampled across five developmental stages: early larvae, late larvae, metamorphosis itself, and early and late juveniles. This technique captures a molecular snapshot of each cell—essentially a readout of which genes are active at a given moment—allowing researchers to classify cells by type and to assess how closely any two cells resemble one another in their identity and function.

The results were striking. When the researchers sorted the cells into twelve broad classes—among them cartilage, immune, and skin cells—they discovered that many larval cells bore a closer molecular resemblance to each other than they did to the adult cells performing the same function. The most dramatic example involved the nervous system: larval neurons were more similar to larval gut cells than they were to the adult neurons they would ultimately give rise to. In other words, the larval neuron’s nearest molecular relative was not the adult neuron, but a completely different tissue in the same larval body. That pattern, repeated across more than half of all the cells analyzed, points to wholesale reprogramming rather than simple continuity.

There were exceptions, and they matter. Muscle cells and mesoderm-derived cells, which form parts of organs, largely retained their identities from larva to juvenile. But the breadth of transformation elsewhere in the body was unlike anything previously documented in a bilaterian during normal development.

Genetic data alone, however, could not rule out one nagging alternative: that the reprogrammed-looking adult cells were actually brand new, born after the original larval cells had perished. To settle the question, Bump used a persistent dye—a lineage tracer—to label larval cells before metamorphosis began, then followed those tagged cells through the transition. The labeled cells survived and persisted into the adult organism, providing direct visual evidence that the larval body is not dismantled and rebuilt from scratch.

“This suggested that cells were not large-scale dying; they were actually being carried over,” Lowe said. “Based on their RNA sequence data, we saw that they had become very different in cell type, which was supportive of the idea that larval cells were being reprogrammed into new fates during metamorphosis, which goes against what most of the field would have predicted.”

The evolutionary context makes the finding all the more tantalizing. Cellular reprogramming has been observed before, but only in circumstances scientists considered unusual: after injury, in animals capable of regenerating lost limbs or entire organs, and in a handful of organisms like sponges and jellyfish during development. Those animals, however, sit far from humans on the evolutionary tree. Schizocardium californicum is different. As a member of the phylum Hemichordata, the acorn worm occupies a branch considered an evolutionary link to vertebrates—the group that includes all mammals, including humans. Finding extensive reprogramming in a normal developmental program of a relative of vertebrates suggests the phenomenon may be far more woven into the fabric of animal development than anyone suspected.

Getting to that discovery was not easy. Schizocardium californicum is not a standard laboratory organism, and the team had to adapt genetic tools and techniques designed for other, better-studied animals to make the analysis possible. But for Lowe’s lab, that difficulty was precisely the point. The group specializes in “non-model” marine organisms, betting that unusual animals will reveal not just their own developmental secrets but broader truths about the evolutionary history of animal life.

The gap they aim to fill is enormous. Most model organisms—mice, zebrafish, fruit flies, and the like—are direct developers: they hatch as miniature versions of their adult selves and grow steadily into maturity. These animals are easier to keep and study in the lab, and they are genetically closer to humans, which is why research has clustered around them. But direct developers are the minority. The majority of animal species pass through a larval stage and undergo metamorphosis, and developmental biology has comparatively little to say about what happens at the cellular level during that passage.

Fittingly, the worm’s own family history helped sharpen the contrast. Schizocardium californicum has a well-studied cousin, Saccoglossus kowalevskii, sometimes called the Virginia acorn worm, which is a direct developer. When the Virginia acorn worm hatches, it already has the worm-like body plan it will keep for life. Its California relative, by contrast, begins existence as a larva that looks nothing like its adult form—and, according to the new study, its interior is transformed as thoroughly as its exterior.

“You can watch this process of metamorphosis and see physically how things radically change, but this morphological transformation is also mirrored by a massive change in cellular components,” Lowe said.

The implications stretch well beyond acorn worms. If cells routinely abandon old identities and adopt new ones during ordinary development in a vertebrate relative, then cellular reprogramming may need to be reclassified from an exotic exception into a mainstream feature of how animals are built. It could also refine how scientists think about regeneration, wound healing, and even the rules that normally keep adult cells locked into their specialized roles.

The research received support from a Chan Zuckerberg Biohub Intercampus Research Award, the National Science Foundation, a Myers Trust Award, and a Haderlie Memorial Award. Additional Stanford co-authors include Laurent Formery, a former postdoctoral scholar, and Lauren Lubeck, a doctoral student in Lowe’s lab, along with researchers affiliated with Baylor College of Medicine, Chan Zuckerberg Biohub in San Francisco, Johns Hopkins University, the Stowers Institute for Medical Research, and the University of California, Berkeley.

For now, the floating head of the Pacific has delivered its verdict: during metamorphosis, the body you inherit may be entirely your own.

The scale of the ambiguity is striking. More than half of the products examined—56 percent—were sold under generic or vague labels such as “shark,” “flake,” or simply “fish and chips,” terms that tell the consumer almost nothing about the animal behind the batter. More concerning still, 35 percent of products were outright mislabelled: the species name advertised on the menu or display did not match the species identified through genetic testing. Even when shops did specify a species—advertising “gummy shark” or “bronze whaler,” for example—the DNA analysis showed that the fillet served did not always contain that species. For a consumer trying to make responsible choices about what they eat, the study suggests that the label on a shark fillet in WA is frequently a poor guide to its true identity.

The methodology behind the findings is a classic application of molecular forensics to the seafood trade. Because cooked and processed fillets lose the morphological features—skin texture, fin shape, colouration—that normally allow species identification, the researchers turned to DNA sequencing to identify each sample genetically. From the 167 cooked fillets, they recovered genetic evidence of 16 different species, 11 of which were sharks. Products sold simply as “shark” turned out to be a composite category, encompassing dusky sharks, gummy sharks and whiskery sharks, among others. The genetic diversity hidden behind that single word illustrates just how interchangeable shark fillets are in the commercial supply chain, both in the eyes of processors and, evidently, in the eyes of vendors.

Lead author Dr Md Robiul Hasan of Murdoch University’s Centre for Sustainable Aquatic Ecosystems, in the School of Environmental and Conservation Sciences, said the findings expose a fundamental transparency problem in how seafood reaches the Australian consumer. “Consumers increasingly want to know what species they are eating, where it comes and whether it has been sourced sustainably,” Dr Hasan explained. “Our results show that this information is often unavailable or inaccurate when it comes to cooked shark products sold in fish and chip shops in WA.” The research was conducted in collaboration with the Western Australian Department of Primary Industries and Regional Development (DPIRD), lending regulatory relevance to the academic findings.

The provenance problem extends beyond species identity to geographic origin. Only 26 percent of the products purchased in the study provided any information about where the seafood came from, leaving the vast majority of customers unable to determine whether their dinner was caught locally or imported from fisheries on the other side of the world. This matters for more than just curiosity: origin information underpins assessments of sustainability, food safety, carbon footprint and support for local fishing communities. Without it, even conscientious consumers have no practical means of aligning their purchases with their values.

Importantly, the researchers are careful not to frame the mislabelling as evidence of deliberate fraud. Many of the shark species identified in the study carry similar commercial value, which makes intentional substitution—a practice documented in seafood fraud investigations elsewhere in the world—a less likely explanation. “Mislabelling may reflect the complexity of commercial fishing, seafood processing and supply chains rather than intentional substitution,” Dr Hasan noted. Shark fillets arriving at processing facilities may be sorted inconsistently, relabelled at multiple points in the chain, or sold under regional trade names that do not correspond to formal species designations. In a supply chain that stretches from fishing vessel to fish and chip fryer, opportunities for species confusion accumulate at every step.

The conservation implications of the findings are nuanced but encouraging in part. While international conservation assessments classify some of the species identified in the study as threatened globally, only a single sample contained a species listed as threatened under Australian conservation assessments. Furthermore, the researchers observed that most of the shark products sampled appeared to originate from species associated with Western Australia’s managed shark fisheries—fisheries operating under state regulatory frameworks with catch controls and monitoring. That distinction between global and national conservation status underscores a recurring theme in seafood sustainability: a species may be in serious decline in one region while remaining sustainably harvested in another, which is precisely why accurate species and origin labelling matters so much for interpreting conservation risk.

One of the most telling findings is regional. When the researchers compared their results with similar DNA-based surveys conducted elsewhere in Australia, they found that the shark species being sold in Western Australia differed from those reported in other states. This regional distinctiveness carries two important lessons. First, it suggests that Australia’s shark seafood market is not monolithic—different states draw on different fisheries, supply networks and trade naming conventions. Second, and more practically, it means that findings from a DNA audit in one jurisdiction cannot simply be extrapolated to another. Effective oversight of seafood labelling, the authors argue, requires regional monitoring and testing tailored to each state’s supply chains, rather than reliance on national averages or one-off national studies.

The study, published in the peer-reviewed journal Marine and Freshwater Research, arrives amid growing international scrutiny of seafood mislabelling. DNA-based audits of fish sold in restaurants, markets and takeaway shops around the world have repeatedly found mislabelling rates ranging from modest single figures to well over half of samples, with sharks, tunas and snappers among the most frequently substituted groups. The WA findings place Australia’s takeaway sector squarely within this global pattern, while also highlighting a distinctive local issue: the deep cultural entrenchment of the term “flake,” a broad Australian trade name historically applied to shark flesh that obscures rather than communicates species identity.

What can be done? The researchers put forward a concrete set of policy recommendations. Chief among them is the mandatory use of standardised fish names for cooked seafood products, so that a fillet advertised as gummy shark must, in fact, be gummy shark. They also recommend origin disclosure requirements, ensuring that consumers can see whether their seafood was caught in Australian waters or imported, and routine DNA auditing as a verification mechanism to keep labelling honest over time. Together, these measures would bring cooked seafood into closer alignment with the labelling standards already expected in many other food sectors, where ingredient identity and provenance are legally protected rather than optional courtesies.

“Clearer labelling would help consumers understand exactly what they are buying and support informed decisions about sustainability, health and seafood provenance,” Dr Hasan said. The study’s wider significance may lie in that link between transparency and sustainability. Shark populations worldwide face pressure from overfishing, and the global shark meat trade has expanded in recent decades as demand for alternative protein sources grows. In that context, every point in the supply chain where species identity is lost or obscured also represents a point where enforcement of fishing regulations, catch quotas and conservation protections becomes harder. DNA testing of a humble fried fillet, it turns out, is not merely a curiosity of molecular biology—it is a tool for tracing how well a multibillion-dollar seafood economy actually works.

For now, the message for Western Australian consumers is a pragmatic one: the “flake” in your paper-wrapped parcel may be any of more than a dozen species, and the name on the board is right only about two-thirds of the time. But the researchers’ findings are less a condemnation of fish and chip shops than a roadmap for fixing a system that currently asks vendors to sell and consumers to buy shark products with almost no reliable information attached. With standardised naming, origin disclosure and genetic auditing, the authors argue, Australia’s takeaway seafood sector could offer transparency that matches both consumer expectations and the realities of marine conservation.

To understand why this matters, it helps to begin with the chemistry. The ocean is Earth’s largest reservoir of iodine, and most of that iodine exists as iodate, a negatively charged ion written chemically as IO₃⁻. Certain bacteria can breathe iodate the way we breathe oxygen, converting it to iodide, I⁻, and harvesting energy in the process. Iodide is far more mobile than iodate, and where it reaches the sea surface it reacts with atmospheric ozone to generate volatile iodine compounds that escape into the air. Once aloft, these species participate in ozone destruction, influence the cycling of mercury, seed the formation of aerosol particles that reflect sunlight, and eventually rain back down onto land, supplying terrestrial ecosystems with iodine. In short, where DIRMs live determines, in a very real sense, how the ocean talks to the atmosphere about iodine.

The classical prediction rested on thermodynamics. On paper, iodate reduction yields more energy than nitrate reduction, so microbial ecologists reasoned that DIRMs should gorge on iodate first, then turn to nitrate. That logic placed them in a narrow ecological window just above the ocean’s oxygen minimum zones—regions where dissolved oxygen has vanished but where denitrifying microbes have not yet exhausted the nitrate. Below that window, nitrate would be scarce; above it, oxygen would suppress anaerobic respiration altogether. Textbook-style reasoning thus confined DIRMs to a thin slice of the ocean.

But field observations from an unexpected quarter—high-iodine groundwater in China—told a stubbornly different story. Professor Junxia Li of China University of Geosciences, the first author of the new paper, had previously isolated a DIRM strain called Azonexus hydrophilus NCP973 from iodine-rich aquifers. Curiously, this organism kept turning up in waters where nitrate had already been depleted, and across high-iodine groundwaters nationwide, iodide concentrations showed a consistent negative correlation with nitrate. “Thermodynamic prediction and field observation were clearly at odds,” Li explains. “We were thus curious to re-examine the ecological niche of these microorganisms.”

The team resolved the conflict in the laboratory. Working with two representative DIRM strains—NCP973 from groundwater and Denitromonas iodatirespirans IR-12, originally isolated from seawater—they supplied cultures with both iodate and nitrate simultaneously. In every experiment, the bacteria reduced nitrate first, and only began consuming iodate once the nitrate had been exhausted. The molecular mechanism emerged from transcriptomic analysis: genes encoding nitrate reductase (narGHI) switched on first, while the iodate reductase genes (idrABP1P2) remained silent until nitrate was gone. Nitrate, it turns out, actively suppresses expression of the iodate reductase machinery, a regulatory strategy analogous to its suppression of perchlorate reductase in other bacteria. Iodate adds a further twist: on its own it imposes oxidative stress that forces the cells into a prolonged lag phase, giving nitrate reduction an unassailable head start.

The implication is striking. If iodate reduction always follows nitrate reduction, then DIRMs should not live above the oxygen minimum zones but inside them, in waters where denitrification is actively consuming nitrate. To test this, the researchers turned to environmental genomics, mining metagenomic and metatranscriptomic datasets from the world’s three major oxygen minimum zones—the Eastern Tropical North Pacific, the Eastern Tropical South Pacific, and the Arabian Sea—along with metagenome-assembled genomes (MAGs) from global OMZs and the Tara Oceans expeditions. The pattern was unambiguous. The idrA gene, the genetic signature of iodate reduction, appeared and was expressed almost exclusively within OMZ depth profiles. Among 962 MAGs recovered from global oxygen minimum zones, 32 carried the full idrABP1P2 gene cluster; among 2,631 Tara Oceans MAGs, only 9 carried it, and every single one came from OMZ samples.

The genomic survey did more than confirm the habitat—it dramatically expanded the known cast of iodate-breathing microbes. The OMZ-dwelling DIRMs belonged predominantly to the candidate phylum SAR324 and to the class Alphaproteobacteria, lineages previously not recognized as iodate reducers. Intriguingly, the SAR324 genomes carrying idrABP1P2 also harbor sulfur oxidation genes, hinting that these organisms may couple the oxidation of sulfide to the reduction of iodate—a chemoautotrophic lifestyle that would let them flourish in the chemically stratified heart of oxygen-depleted waters. To verify that the newly discovered genes actually do what they appear to do, the team expressed two representative idrABP1P2 sequences heterologously in laboratory hosts and confirmed iodate-reducing activity.

Rewriting the map of DIRM habitats is more than an exercise in microbial cartography; it changes projections of the marine iodine cycle in a warming world. Global warming lowers the solubility of oxygen in seawater and intensifies ocean stratification, both of which encourage the expansion of oxygen minimum zones. Over the past six decades, the global area covered by OMZs has grown from roughly 5 percent of the ocean to 14 percent, and the trend is expected to continue. Every square kilometer of new oxygen-depleted water is, under the revised model, potential habitat for DIRMs—more microbes converting iodate to iodide, and more iodide available for transport to the surface ocean by circulation. Model simulations cited in the study indicate that a 1 percent increase in global sea-surface iodide concentration produces approximately a 0.7 percent rise in oceanic iodine emissions to the atmosphere.

Those emissions matter on several fronts. Volatile iodine compounds destroy tropospheric ozone, alter the oxidation chemistry of the marine boundary layer, contribute to the formation of new aerosol particles that influence cloud cover and climate, and govern the atmospheric fate of mercury, a potent neurotoxin. They also deliver iodine back to land, where adequate dietary intake prevents goiter and other thyroid disorders that still afflict populations in iodine-poor regions. A microbe’s preferred address in the water column thus ripples outward to human nutrition and planetary climate chemistry alike.

The study also offers a methodological lesson that extends well beyond iodine. Thermodynamic calculations, however elegant, describe what is energetically possible, not what regulatory networks actually permit. The discovery that nitrate represses idrABP1P2 expression—inverting the presumed order of substrate use—illustrates how gene regulation can override energy-yield logic in shaping microbial niches. For biogeochemists accustomed to predicting microbial distributions from redox potentials alone, the message is clear: transcriptomics and genome-resolved surveys of the actual environment must carry at least equal weight. Ecological niches, this work reminds us, are written in regulatory circuits as much as in reaction energetics.

Li and colleagues argue that their findings should now be integrated into marine iodine biogeochemical models. Doing so would sharpen predictions of how oceanic iodine emissions will respond as oxygen minimum zones continue to expand through the coming century. Given that iodine touches everything from stratospheric ozone recovery to cloud droplet formation to human endocrine health, the stakes of getting this cycle right are considerable. The humble iodate-breathing bacteria of the deep, once relegated to a theoretical sliver of the ocean, now appear to command a far larger territory—and their expanding domain may help determine the atmospheric chemistry of a warming planet. “Given the importance of iodine for human and environmental health,” Li notes, “integrating this pathway into marine iodine biogeochemical models will improve our capability of understanding and predicting the future changes in oceanic iodine emissions.”

The findings, published in the Zoological Journal of the Linnean Society, represent a triumph of what researchers are calling “museomics” — the extraction and analysis of DNA from century-old museum specimens. By combining this historical genetic material with freshly collected specimens and cutting-edge imaging techniques, the team reconstructed the family tree of branching syllid worms and solved a puzzle that has vexed zoologists since the 1870s.

Branching worms belong to a group of annelids — segmented worms related to earthworms and leeches — but their anatomy defies nearly everything we associate with worm body plans. While a typical worm grows by adding length, these creatures grow by splitting. A single head gives rise to a body that repeatedly bifurcates into a sprawling network of posterior branches, creating a living bush whose crown can spread throughout the interior of a sponge host. Scientists believe this bizarre architecture is an adaptation to their symbiotic lifestyle: by branching through the sponge’s internal canals, the worm can occupy its host thoroughly without ever exposing itself to the open water.

Until now, only three species of branching worms had ever been described. The most famous, Ramisyllis multicaudata, was discovered in Australia in 2006 and announced to the world in 2012, decades after the first branching worm, Syllis ramosa, was found inside a sponge collected in the Philippines during the HMS Challenger expedition. That specimen, collected in 1875, launched the mystery: how could such strange animals exist, how many were there, and how on earth did their branching bodies evolve?

The new study answers at least some of these questions with unprecedented detail. The research team, led by Professor Maria Teresa Aguado, Scientific Curator of the Biodiversity Museum at Göttingen University, assembled material from an extraordinary range of sources: worms collected from shallow reefs and deep-sea expeditions across the Indo-Pacific, the Red Sea and New Zealand, alongside fragile historical specimens preserved in museum collections, some of which date back to 1914. Sponge specimens — the hosts in which these worms live hidden — were examined non-destructively using microCT imaging, a technique that uses X-rays to build detailed three-dimensional models of an object’s interior without cutting it open. In several cases, researchers could confirm the presence of branching worms coiled inside sponges that had been collected over a century ago.

“Our findings would have been impossible without access to valuable museum specimens,” said Professor Aguado. “These collections are not just repositories of old samples, but resources that continue to generate new discoveries.”

The genetic and anatomical analyses revealed thirteen distinct forms of branching worms, at least ten of which are likely new to science. Equally significant was what the DNA told the researchers about how these animals relate to one another. Branching worms, it turns out, are not a single loosely related group of oddities — they form a coherent evolutionary lineage that split into two major branches, rather like the bodies of the worms themselves. One lineage retains the genus name Ramisyllis. The other has been assigned to an entirely new genus, Cladosyllis, described in the same study.

The deeper evolutionary story is perhaps the most striking of all. By comparing genomic data across the group, the researchers demonstrated that all branching worms descend from a single common ancestor — meaning that their extraordinary tree-like body architecture evolved only once in the history of life. That single evolutionary event, apparently rare enough that no other animal group has replicated it so dramatically, gave rise to a worldwide radiation of species. Yet despite their shared origin, the two major lineages produce their branches in fundamentally different ways, a discovery that highlights the developmental flexibility evolution can exploit. The same branching outcome, it seems, can be reached through different embryological and growth mechanisms — a finding with implications for how biologists understand the evolution of complex body plans across the animal kingdom.

The distribution of the newly discovered species tells its own story. Branching worms were found across an enormous geographic range: the tropical and subtropical waters of the Indo-Pacific, the Red Sea, and the waters around New Zealand. Some species live in shallow-water sponges accessible to divers, while others were found inside deep-sea glass sponges at depths of up to 1,000 metres — creatures like Crateromorpha meyeri, a delicate lattice of silica whose translucent body can reveal the semi-transparent threads of a resident Cladosyllis worm at its main opening. The researchers believe this close association with specific sponge hosts has been a major engine of diversification. As different worm lineages adapted to different sponge species — and to vastly different environments, from sunlit reefs to the abyssal dark — they diverged into the distinct species now being described.

“For over a century, the rare branching worms were all classified as a single widespread species,” said Dr Guillermo Ponz Segrelles, a co-author of the study. “However, we have shown that – like their bodies – their family tree has many branches, each closely associated with its own sponge species and located in specific areas.”

The revelation that a “single” species actually encompasses a dozen or more distinct lineages is a familiar refrain in modern taxonomy, but rarely has the hidden diversity been so dramatic — or so physically strange. It also underscores the critical role that natural history museums play in twenty-first century science. Specimens collected generations ago, preserved in alcohol and archived in institutional drawers, have become a genetic treasure trove. Techniques that were unthinkable when these worms were first collected — high-throughput DNA sequencing, microCT scanning, computational phylogenetics — can now coax secrets from material that was once considered little more than taxonomic paperwork.

“Some of the sponges hiding these worms came from the Senckenberg collection and date back to 1914,” noted Dr Ekin Tilic of the Senckenberg Research Institute and Natural History Museum Frankfurt, another co-author of the study. “We were able to reveal the worms without damaging the sponges by using microCT imaging – modern techniques are helping us rediscover hidden biodiversity in centuries-old collections.”

The study also raises questions that will keep researchers busy for years. How does a worm with a single head coordinate a body with dozens of posterior ends? Earlier work on Ramisyllis multicaudata suggested its nervous system branches in parallel with its body, and that the worm reproduces in a peculiar fashion — releasing specialized stolons bearing gametes that swim away from the branch tips. Whether the newly described Cladosyllis species share these traits, and how their different branching mechanisms develop, remains to be explored. The researchers suggest that branching worms could become an important model system for studying the evolution of complex body forms, the biology of symbiosis, and the maintenance of biodiversity in marine ecosystems.

There is also a conservation dimension to the discovery. Because branching worms depend on sponge hosts, and because sponges — particularly the slow-growing glass sponges of the deep sea — are vulnerable to trawling, warming oceans and habitat disturbance, the fate of these worms is inseparable from the fate of their hosts. The full extent of their diversity is likely far greater than the current tally suggests. The deep sea remains one of the least explored environments on Earth, and if a single sponge species collected in 1875 could hide a branch of the tree of life for nearly 150 years, the ocean’s unexplored sponges may harbor many more.

What began as a curiosity preserved in the collections of the Challenger expedition has become a window into the creative power of evolution. One head, one ancestor, one evolutionary innovation — and from it, a tree of worms branching through the oceans, some of them hidden for over a century in the quiet halls of the world’s museums, waiting for the right tools to reveal them.

The timing of the analysis matters. The meetings in Nice took place under the umbrella of the United Nations Decade of Ocean Science for Sustainable Development, an ambitious international initiative explicitly designed to produce ocean science that fosters equity and sustainability. The “One Ocean” framing that unified participants in Nice was intended to be just that, a unifying banner under which nations, scientific institutions, Indigenous communities, industry, and civil society could coordinate their efforts on behalf of a shared planetary resource. Yet according to the researchers, the very act of framing the ocean as one, and the science that describes it as universal, raises difficult questions about whose knowledge counts, whose ways of knowing the ocean are included, and what happens to knowledges that do not fit the universal template.

The authors bring considerable methodological firepower to the question. They draw on their prior scholarship on new data technologies and ocean governance, combined with participant observation at both the congress and the conference. This ethnographic approach allowed them to watch, in real time, how universalist claims about ocean knowledge were made, negotiated, and occasionally disrupted on the international stage. Their findings, distilled into the new paper, center on three sites where the tension between universalism and difference was most visible: the tendency to render the oceans technical rather than political, the persistent and even constitutive role of gaps in ocean data, and the existence of sciences and knowledges that are fundamentally incommensurable with One Ocean Science’s universalist ambitions.

The first of these sites, the technical rendering of the ocean, is perhaps the most recognizable to anyone following ocean science today. At the Nice meetings, the authors observed a strong propensity to present the ocean as a domain that can be fully mapped, measured, modeled, and monitored. New data technologies, including sensor networks, autonomous platforms, digital twins, and large-scale data integration efforts, promise what the researchers describe as “complete” ocean renderings, comprehensive digital portraits of the sea that ostensibly leave nothing out. The appeal is obvious: a fully quantified ocean seems to offer a neutral, objective basis for management and conservation decisions. But the researchers caution that this technical framing does political work. When the ocean is rendered primarily as a technical object, the political questions, who benefits from ocean resources, who bears the costs of conservation, whose voices shape governance, recede into the background, treated as settled or secondary to the business of data collection.

This is not an argument against data. The authors acknowledge the genuine scientific value of new observation technologies and the enormous effort behind them. Their point is subtler and more provocative: even the most ambitious data infrastructures embody choices about what to measure, how to categorize, and which questions matter. A “complete” ocean rendering is complete only relative to a particular framework of what counts as knowledge of the ocean. Frameworks are built by people, institutions, and funding structures, all of which carry histories, interests, and blind spots. Rendering the ocean technical, the researchers suggest, can obscure rather than resolve the politics woven through every dataset.

The second site of tension is the persistence of gaps. Ocean science has long been characterized by the fact that most of the ocean remains under-observed, under-sampled, or entirely unmeasured. The researchers found that gaps functioned as a defining feature of One Ocean Science at the Nice meetings, invoked repeatedly as a rationale for new expeditions, new funding, and new technologies. The gap talk is productive for science advocacy, precisely because it generates urgency. Yet the authors identify a deeper implication: gaps are not merely temporary deficits to be closed by better instrumentation. Some gaps reflect structural inequalities in the global scientific enterprise, including unequal access to research vessels, satellite data, computational capacity, and publication venues. A universalist science that treats all gaps as the same kind of problem risks overlooking the fact that some gaps are the direct product of historical and political asymmetries.

The third and arguably most challenging site concerns knowledges that simply cannot be rendered commensurable with One Ocean Science universalism. The ocean is known in many ways: by coastal communities whose intergenerational knowledge is embedded in practice and place, by fishers whose livelihoods depend on intimate familiarity with local waters, by Indigenous peoples for whom the ocean is woven into identity, law, and cosmology, and by formal disciplines from physical oceanography to marine ecology. The researchers observed that while the Nice meetings made space for difference rhetorically, celebrating inclusivity and pluralism, the underlying epistemological infrastructure of the events, the session formats, the data standards, the metrics of success, remained organized around a particular model of scientific knowledge. Knowledges that resist translation into that model, that cannot be converted into harmonized datasets or policy-ready indicators, remained marginal. Universalism, in other words, has edges, and the people and practices that fall outside them are not accidental casualties but structural features of how One Ocean Science operates.

Here the paper delivers its most counterintuitive claim. The researchers argue that One Ocean Science universalism is not undermined by difference; it is constituted through it. The very coherence of the “one ocean” narrative depends on heterogeneous scientific practices, diverse political interests, and uneven capacities being brought together, translated, and synchronized. Universalism is not a description of a pre-existing unity but an achievement, continuously produced through work that manages, absorbs, or excludes difference. This insight reframes the challenge of equitable ocean science. The task is not to eliminate difference in the name of unity, nor to reject universal science, but to hold space for difference within One Ocean, allowing plurality to shape the scientific enterprise rather than treating it as residue to be cleaned up.

The practical implications are significant. The authors argue that manifesting the goals of the UN Decade of Ocean Science, particularly the commitments to equity and sustainability, will require interrogating and redesigning what they call the epistemological infrastructure of One Ocean Science. That phrase encompasses the conventions, standards, platforms, and institutional routines through which ocean knowledge is produced, validated, and circulated. Designing pluralism into that infrastructure, making it structural rather than residual, could mean rethinking data governance so that community-held knowledge is protected rather than extracted, restructuring conference and decision-making formats so that non-dominant knowledge holders are participants rather than guests, and building evaluation criteria that recognize multiple forms of rigor and relevance.

The research also speaks to a broader conversation in science and technology studies about the politics of knowledge production in global environmental governance. Ocean governance is entering a decisive period, with a new international treaty on marine biodiversity in areas beyond national jurisdiction moving toward implementation, deep-sea mining debates intensifying, and climate change reshaping ocean chemistry and circulation at pace. In this context, questions about whose knowledge informs governance are not academic. Decisions made on the basis of particular knowledge frameworks will distribute benefits and burdens across nations, communities, and generations. The researchers’ analysis suggests that acknowledging the politics of knowledge is not an obstacle to effective ocean governance but a precondition for legitimate and durable outcomes.

The team behind the paper is notably international and interdisciplinary, spanning geography departments at the University of North Carolina at Chapel Hill, the University of Guelph, Durham University, and Université Côte d’Azur, along with institutions including KTH Royal Institute of Technology in Stockholm, the Helmholtz Institute for Functional Marine Biodiversity at the University of Oldenburg, the Alfred Wegener Institute, the Duke University Marine Lab, and the University of Vienna. The work was supported by funders including the U.S. National Science Foundation, the Helmholtz Association, the French CNRS, and the European Research Council, none of which, the authors note, played a role in study design or interpretation. Corresponding author Oscar Hartman Davies and his colleagues conducted their fieldwork amid the bustle of conference halls and negotiation rooms, capturing a moment when the world’s ocean community assembled around a shared banner and, in doing so, revealed the seams beneath it.

The message the researchers hope will travel beyond Nice is ultimately constructive. One Ocean Science has achieved remarkable things, mobilizing resources, attention, and cooperation on a planetary scale. But if the Decade of Ocean Science is to deliver science that is genuinely equitable and sustainable, the ocean community must make room for the many knowledges of the ocean, not as decorative additions to a universal edifice, but as constitutive parts of it. The one ocean, the authors remind us, has always been known in many ways. The future of ocean science depends on whether its institutions are willing to know that too.

The study, led by Professor Emeritus Yuji Oshima of Kyushu University’s Faculty of Agriculture and published in the journal Marine Pollution Bulletin, combined laboratory exposure experiments with multi-omics analysis—a suite of techniques that simultaneously examines gene expression and microbial communities to build a comprehensive picture of an organism’s internal state. The choice of the wharf roach as a study subject was deliberate. Previous fieldwork by the team had shown that these animals chew up expanded polystyrene and excrete it as much smaller fragments, raising the possibility that wharf roaches act as unwitting factories converting bulky foam debris into microplastics. But while that ecological role was becoming clear, the biological cost to the animals themselves remained an open question. “We wanted to find out whether swallowing foam comes at a biological cost for these shoreline scavengers,” Oshima explained in the announcement of the findings.

The experimental design was elegantly simple. Field-collected wharf roaches were divided into two groups: one received nothing but pieces of a commercially available polystyrene foam board for one week, while a control group received no food at all. The first measure the researchers examined was survival. Foam-fed wharf roaches lived for an average of 27.8 days, compared with 31.6 days for the starved controls—a difference of 3.8 days that, despite its direction, did not reach statistical significance. On the surface, then, the animals appeared remarkably resilient. The plastic did not visibly shorten their lives under laboratory conditions, and the feeding animals looked outwardly healthy. Had the researchers stopped there, the story might have ended with a reassuring conclusion: eating polystyrene foam, at least for a week, does not kill a wharf roach.

But the multi-omics analysis told a far more intricate story. When the researchers examined gene expression in the digestive tracts of the EPS-fed roaches, they found a suite of changes centered on the animal’s chemical defense machinery. Three families of detoxification enzymes were expressed at elevated levels: cytochrome P450 enzymes, UDP-glucuronosyltransferases, and sulfotransferases. These enzyme families are the workhorses of xenobiotic metabolism, the biochemical system by which organisms neutralize foreign compounds. Cytochrome P450 enzymes typically handle the initial chemical modification of harmful molecules, while UDP-glucuronosyltransferases and sulfotransferases carry out the subsequent phase, conjugating those modified compounds with molecules that make them water-soluble and easier to excrete. Together, the elevated expression observed in the foam-fed roaches represents both major phases of the classical detoxification pathway firing in response to the ingested plastic.

The pattern of gene expression changes extended beyond detoxification. A gene involved in DNA repair also showed higher expression in the EPS-fed animals, a change that may reflect cellular stress or damage to genetic material, though the precise trigger remains to be determined. At the same time, several genes encoding digestive enzymes were expressed at lower levels in the foam-fed group, including one involved in breaking down plant fibers. This downregulation of digestive machinery makes intuitive sense: an animal consuming a nutritionally barren substrate like polystyrene foam may have little use for the full complement of enzymes designed to extract energy from genuine food. The gut, in effect, was reprogramming itself around a diet that provides bulk but no nourishment.

Perhaps the most surprising findings came from the microbiome analysis. The researchers expected that a radical dietary shift from normal scavenged organic matter to pure plastic foam would restructure the community of microbes living in the roaches’ guts. It did not. Microbial diversity within individual animals and the overall composition of gut communities were essentially unchanged across the four microbial domains the team examined. The core microbial ecosystem of the wharf roach gut, it seems, is stable enough to weather a week of plastic dining without wholesale disruption. But beneath that stable surface, the researchers detected something intriguing: several rare organisms appeared exclusively in the plastic-fed animals. Three archaeal taxa, including Methanospirillum, a genus of methane-producing archaea, along with one family of bacteriophages, were detected in all three of the EPS-fed specimens examined but in none of the controls.

The appearance of methane-producing archaea in the guts of plastic-fed roaches is a detail with potentially broad implications. Methanogens are typically associated with anaerobic, fermentative environments, and their presence suggests that the digestion of polystyrene foam may alter the chemical microenvironment of the gut in ways that favor these specialized microbes. Similarly, the appearance of specific bacteriophages—viruses that infect bacteria—only in the foam-fed group hints at subtle shifts in microbial dynamics that standard diversity metrics failed to capture. Rare taxa, the researchers note, can serve as sensitive indicators of environmental change, and their group-specific patterns here suggest that even a microbiome that looks statistically “unchanged” may be harboring meaningful shifts beneath the surface.

For Oshima, the discrepancy between the animals’ outward health and their internal molecular activity is the central lesson of the study. “The foam-fed animals appeared healthy and had lifespans similar to those of the controls; however, their guts showed differences in the expression of genes involved in chemical defense,” he noted. “This tells us ‘no visible harm’ does not necessarily mean ‘no biological effect.'” That distinction matters far beyond the wharf roach. Environmental toxicology has long grappled with the problem of sublethal effects—changes in an organism’s physiology that do not cause obvious illness or death but may nevertheless impose energetic costs, reduce reproductive success, or compromise resilience to other stressors. Running a detoxification system at elevated capacity, repairing DNA, and remodeling digestive machinery all consume energy and resources that the animal cannot then spend on growth, reproduction, or immune defense. Whether the gene expression changes observed here translate into functional consequences for wharf roach populations in the wild is a question the researchers say demands further work.

The broader context of the study is the escalating global problem of expanded polystyrene pollution. EPS is inexpensive, lightweight, and easy to manufacture, which has made it a staple material in the fishing and packaging industries. Those same properties—low density and high buoyancy—make it an exceptionally persistent marine pollutant, capable of drifting across vast ocean distances before accumulating on beaches and shorelines. Once in the environment, EPS fragments readily into smaller and smaller pieces, seeding coastal habitats with microplastics. The Kyushu team’s earlier fieldwork added a troubling dimension to this picture: wharf roaches themselves may actively accelerate this fragmentation, chewing foam into particles small enough to enter food webs at multiple trophic levels. Shoreline scavengers, in other words, are not merely victims of plastic pollution—they may be agents in its dispersal and transformation.

The findings arrive at a moment of growing scientific concern about the biological effects of microplastics on invertebrates, which form the foundation of many marine food webs. Wharf roaches occupy a particularly valuable position as a study organism and as a potential sentinel species for coastal plastic pollution. Because they live at the land-sea interface, consume whatever washes ashore, and are abundant and easy to collect, changes in their physiology and microbiomes could serve as an early warning system for the biological impacts of plastic accumulating along the world’s coastlines. A sentinel that displays measurable molecular responses to a common pollutant, even without overt signs of harm, offers researchers a sensitive instrument for tracking environmental contamination.

The research team, which included Seokhyun Lee, Hirokuni Miyamoto, Yuki Takai, Wataru Suda, Hiroshi Ohno, and Yohei Shimasaki alongside Oshima, frames its results as yet another argument for managing expanded polystyrene waste more responsibly. The message is twofold. First, improving the design, recovery, and disposal of EPS products would reduce the volume of foam entering marine environments in the first place. Second, prioritizing coastal cleanup efforts remains essential, because stranded foam does not simply sit inertly on the beach—it is eaten, fragmented, and biologically transformed by the very animals that help keep shorelines clean. The wharf roach, it turns out, pays a molecular price for its cleanup work. Whether that price compounds across generations and ecosystems is the question this research now opens.

News Publication Date: 4-Sep-2026

Web References: Kyushu University Faculty of Agriculture; Kyushu University

References: Lee, S., Miyamoto, H., Takai, Y., Suda, W., Ohno, H., Shimasaki, Y., & Oshima, Y. (2026). Changes in dysbiosis and gene expression in the gut of wharf roach (Ligia spp.) fed with expanded polystyrene. Marine Pollution Bulletin. https://doi.org/10.1016/j.marpolbul.2026.120200

The technique, known as applied nucleation, is well established in the restoration of terrestrial forests. Rather than planting an entire degraded landscape, practitioners establish small clusters of vegetation and allow natural propagule dispersal, facilitation and succession to do the heavy lifting. The approach is particularly valuable where seed or spore sources are absent, dispersal capacity is limited, or the environmental conditions required for early establishment depend on the presence of a canopy. Translating this logic to the ocean, however, had never been tested at this temporal and spatial scale. The new study provides the most comprehensive evidence to date that the concept can work underwater, while also revealing the ecological variables that determine success and failure.

The study organism, Phyllospora comosa, commonly called crayweed, is a fucoid seaweed that once formed dense forests along the shallow subtidal reefs of temperate Australia, from the low tide mark to roughly six metres depth. These forests support distinct assemblages of epifauna and are associated with higher abundances of commercially important species such as rock lobster and abalone. During the 1980s, crayweed disappeared entirely from the Sydney metropolitan coastline, a loss that coincided with the direct discharge of sewage onto the city’s beaches and reefs. When onshore sewage outfalls were decommissioned in the early 1990s and replaced with deepwater ocean outfalls, water quality improved rapidly, yet crayweed failed to return naturally. Genetic evidence suggested that populations along the coast remain connected and that dispersal over long distances is theoretically possible, which pointed researchers toward recruitment limitation as the most likely barrier: no nearby reproductive adults, and therefore no local supply of propagules, meant the species could not reclaim its former habitat despite suitable water chemistry.

Operation Crayweed, the restoration programme at the heart of the new analysis, began in 2012 at Long Bay. There, scientists established a patch of approximately 20 square metres by attaching around 400 reproductive adult crayweed, collected from extant forests north and south of Sydney, onto mats drilled into the reef at densities of roughly 15 to 20 individuals per square metre, mirroring the densities observed in natural forests. Subsequent sites used smaller clusters of mats covering 12 to 20 square metres, typically along about five to ten metres of coastline. In total, transplanting events were carried out at 16 sites between 2012 and 2024, with an average of approximately 475 reproductive adults transplanted per site. The study compiled monitoring data from 14 of these sites across 13 years, combining transect surveys, snorkel-based GPS mapping, underwater visual census, stereo-video fish surveys, light and temperature loggers, and accelerometer-based measurements of wave motion.

The results from the earliest site were striking. Within 36 months of transplantation at Long Bay, crayweed individuals had established up to 43 metres from the initial patch, with significantly more recruits found closer to the mat than farther away. Critically, the re-established individuals were not stunted remnants: their lengths reached the ranges documented in extant populations within three years, and the proportion of reproductive adults climbed from 50 percent at 16 months to 88 percent at 36 months. In other words, the transplanted patch did not merely persist; it seeded a new generation capable of reproducing and expanding on its own, exactly the trajectory that applied nucleation is designed to set in motion.

Across the broader programme, the picture was more nuanced. Crayweed established successfully at six of the 14 transplanted sites, a success rate of roughly 43 percent. Where establishment occurred, the re-established populations extended up to 388 metres from their initial patches after 12 years, and the area enclosed by the outer edges of the recovered population increased by approximately 96,380 percent relative to the original planted footprint. Densities of individuals within recovered patches ranged from 1 to 34 per square metre, though the site-wide average of around 2 individuals per square metre remained below the 14 per square metre recorded in reference populations outside Sydney, indicating that full recovery remains a work in progress.

The statistical modelling behind the study identified several factors that separated successful sites from failures. In the short term, recruitment measured nine months after transplantation was positively associated with the survival of the transplanted adults, which serve as both a source of gametes and a canopy sheltering recruits from excessive light and physical disturbance. Recruit length, meanwhile, was negatively associated with grazing damage on the transplants, suggesting that herbivory, even when not correlated with the raw abundance of urchins, snails or herbivorous fish, imposes real costs on early life stages. Over the longer term, a clear pattern emerged around canopy: crayweed expanding away from the original patch was found far more often than expected adjacent to other canopy-forming seaweeds, particularly the kelp Ecklonia radiata and Sargassum species, rather than on bare rock or turf-dominated substrate. This facilitation effect implies that site selection models should explicitly account for the presence of neighbouring canopy-formers, which moderate light, reduce thallus scour and possibly suppress herbivore access.

Timing also mattered. Most sites where crayweed successfully established had been transplanted between April and November, particularly during the austral winter, whereas summer transplanting attempts largely failed. This aligns with the reproductive phenology of the species: gamete release and maturity peak in winter and decline sharply in summer. When the analysis was restricted to winter transplanting events, the positive relationships between the number of transplanting events and both the extent and area of re-established crayweed became significantly stronger. Repeated transplanting, or reinforcement, also improved outcomes, echoing findings from terrestrial restoration where repeated planting buffers populations against unpredictable disturbances such as storms or grazing pulses. The researchers note that severe flooding at Kurnell in 2022 likely wiped out a successfully recruited population there, underscoring that even well-chosen sites remain vulnerable to extreme events, pollution sensitivity and ocean warming.

The team also translated their expansion rates into projections for meeting global restoration ambitions. The Kunming–Montreal Global Biodiversity Framework has prompted a target of protecting three million and restoring one million hectares of marine forests by 2040, yet only about 15,000 hectares have been restored to date worldwide. Applying their measured expansion rates to Sydney’s 56,713 metres of exposed rocky reef coastline, the researchers calculated that establishing 30 simultaneous new sites would allow crayweed to reclaim 30 percent of suitable degraded habitat in approximately 47 years, at an estimated cost of USD 95,788, based on a linear restoration cost of about USD 5.63 per metre that includes materials, transport and personnel. If only sites where crayweed has actually established are considered, the timeframe shrinks to 21 years, and excluding urchin barrens, which occupy roughly 29 percent of Sydney’s reefs and would require additional interventions such as urchin culling, would further alter the calculus. The projections do not account for potential climate-driven losses, although the spread rates inherently incorporate periodic grazing and storm-related setbacks.

The authors caution that a fundamental question remains: whether re-established crayweed forests, even at target scales, deliver the full suite of ecological functions, biodiversity and ecosystem services characteristic of extant forests. Answering this will require monitoring over 15 to 20 years, incorporating metrics of biodiversity, function and service provision, particularly as climate change reshapes the suitability of restoration sites. Future-proofing strategies under consideration include selecting thermally resilient genotypes, reinforcing restored populations, and using spatially explicit climate models to guide site selection. Meanwhile, modified green gravel techniques, in which juvenile seaweeds are seeded onto small gravel or rock substrates for outplanting, are being developed for crayweed to push scalability further.

What the study ultimately demonstrates is that small, strategic interventions can leverage natural ecological processes to achieve restoration outcomes at scales far exceeding the initial investment. Twelve- to twenty-square-metre patches of transplanted adults have, over a decade, given rise to expanding marine forests along one of Australia’s most urbanised coastlines. The findings offer a practical blueprint for practitioners worldwide confronting the decline of kelp and fucoid forests, roughly half of which have degraded over the past 50 years due to overfishing, pollution and ocean warming. From little things, as the paper’s title suggests, big things can indeed grow, provided that the biology of the target species, the ecology of the site and the realities of herbivory, seasonality and disturbance are woven into the restoration design from the outset.

The discovery emerged from a research effort led by scientists at Sun Yat-sen University (SYSU) and the Southern Marine Science and Engineering Guangdong Laboratory in Zhuhai. Using modern deep-sea diving vehicles, including human-occupied vehicles and remotely operated vehicles, the team filmed deep-sea organisms in their natural habitat across three areas of the northern South China Sea. Among the animals they documented were three species of peristediid fishes, commonly called armored searobins: Scalicus engyceros, Paraheminodus murrayi and Peristedion liorhynchus. The footage captured not just still portraits of these animals but extended records of their benthic behaviors, allowing the researchers to describe, for the first time, how the fishes move and forage while alive on the seafloor.

Han Tian, the paper’s primary author and a doctoral researcher in the School of Marine Sciences at Sun Yat-sen University, described the finding as a paradigm-shifting revelation. S. engyceros belongs to a group of fishes characterized by highly specialized free pectoral-fin rays, finger-like appendages that extend from the pectoral fins and make contact with the seafloor. Scientists have long suspected that searobins use these rays to walk, but until now the behavior had not been directly confirmed with in situ footage. The species has even attracted popular attention for its appearance, which combines features of fish and shrimp so strikingly that U.S. local media have dubbed it a “fish-prawn hybrid.”

The walking itself proved more versatile than anyone anticipated. Searobins were first described by the zoologist Albert Günther in 1872, yet in the more than 150 years since, no one had predicted that these fishes could walk both sideways and backward. Locomotion in reverse has never been observed in other fishes, making the armored searobin’s gait a genuine first for the field. The finding illustrates how much basic behavioral information remains locked in the deep sea, inaccessible until organisms can be observed alive in their own environment rather than studied from preserved specimens hauled up in trawls.

The searobin’s body plan reveals a suite of adaptations that appear tailored to life on soft seafloor sediment. Beyond the walking rays, the fish sports outward-extending barbels that Tian compared to a farmer’s rake. While the structure may look awkward for getting around, it serves a foraging function: the barbels allow the searobin to sense and probe potential prey buried in the seafloor sediment, and even to dig into the surface sediment itself in search of food. Meanwhile, the animal’s main pectoral fins have evolved into flat, round plates, which appear to improve balance both during walking and while swimming. When threatened, the fish can deploy its shrimp-like fin rays and tail to make explosive, jerky leaps reminiscent of a startled shrimp, a rapid escape behavior that complements its slower, deliberate walking.

The filmed observations also yielded unexpected insights into the searobin’s sensory world. The fish’s large eyes could not resolve the approaching vehicle itself, indicating that its visual acuity is limited at close range. Yet when the vehicle’s light beam swept near, the animal rolled its eyes toward the light. This response suggests that the searobin’s eyes have retained some sensitivity to light, which, according to the researchers, may reflect a light history during the species’ life cycles. Many deep-sea fishes begin life in shallower, sunlit waters before descending to the abyss, and the eye-rolling behavior hints at vestiges of that developmental or evolutionary past, though the team presents the observation as a suggestion rather than a settled conclusion.

The fact that three different armored searobin species were observed at different locations in the South China Sea raised another question for the team. The distribution of these similar fishes across separate sites suggests, in the researchers’ view, the hypothesis that local specialization forces may be at work, shaping each population’s adaptations to its particular patch of seafloor. Testing that hypothesis will require further observation and comparison, but it points toward a research program that treats deep-sea fishes not as static curiosities but as dynamically evolving lineages responding to the distinct conditions of their habitats.

The methodological significance of the study lies in its approach. Classified as an observational study, the work rests on in situ filming rather than laboratory experimentation or dissection. Advances in deep-sea vehicle technology, including human-occupied vehicles and remotely operated vehicles, have transformed what is possible for deep-sea biology. Where earlier generations of scientists had to infer behavior from the anatomy of dead specimens, today’s researchers can watch living animals interacting with their environment. Tian emphasized that modern diving vehicles allow not only the discovery of new species but also add a new in situ, functional dimension to the study of species that were previously known only from preserved morphology. Observing deep-sea animals alive in their surroundings, the team argues, provides direct insight into adaptation and evolution that museum specimens alone cannot supply.

The broader context makes the finding all the more compelling. The deep ocean is one of the most inhospitable places on Earth, characterized by crushing pressure, perpetual darkness and scarce food, yet many species have managed to survive there, often evolving specialized body structures and behaviors in the process. Armored searobins represent one such evolutionary experiment: a lineage that has repurposed its pectoral fins into a hybrid toolkit serving walking, balance, foraging and escape. The combination of rake-like barbels for probing sediment, plate-like fins for stability, free rays for stepping across the bottom and shrimp-style leaps for emergency escapes suggests a tightly integrated adaptive complex rather than a single isolated trait.

Looking forward, the team intends to push beyond this initial documentation. Tian said the researchers aim to explore the co-evolution of the fish’s unique rake-shaped appendages, its walking locomotion and its deep-sea foraging strategies, with the goal of establishing a new framework for understanding benthic fish adaptive evolution. Such a framework could reshape scientific perceptions of deep-sea biodiversity and illuminate how extreme marine environments drive the emergence of biological structures and behaviors seen nowhere else in the world’s aquatic ecosystems. In the team’s view, these walking fish harbor far more novel behavioral and evolutionary adaptations than previously assumed.

The research was a collaborative effort. Alongside Tian, contributors included Wei Xie, Mingting Li and Kedong Yin, affiliated with the School of Marine Sciences and the Guangdong Provincial Key Laboratory of Marine Resources and Coastal Engineering at Sun Yat-sen University, and with the Southern Marine Science and Engineering Guangdong Laboratory in Zhuhai, China. The paper, titled “Walking Fish,” carries the DOI 10.34133/olar.0172, and the authors reported no conflicts of interest. As with any observational study, the findings describe what the cameras captured in a limited number of encounters and locations; the mechanisms underlying the backward gait, the developmental basis of the fish’s light sensitivity and the role of local specialization remain open questions for future work. Still, the footage marks a milestone: after a century and a half in the scientific record, the armored searobin has finally been seen doing what its anatomy always hinted at, walking, in every direction, across the dark floor of the South China Sea.