Mesophotic ecosystems occupy the dimly lit zone of the ocean, roughly between 30 and 150 meters in many subtropical regions, where light is too scarce for most photosynthetic reef builders but where life still flourishes in remarkable abundance. In many oceanic archipelagos, these communities are dominated by black corals of the order Antipatharia, a group of more than 300 species characterized by proteinaceous skeletons, an almost universal lack of symbiotic algae, slow growth and long lifespans. Whip black corals are a distinctive form within this group: they grow as a single unbranched stem that can extend several meters in length, spiraling and swaying in areas of strong current. Where they aggregate in high densities, they form what ecologists call Marine Animal Forests, three-dimensional structures built by suspension-feeding animals that alter water flow, trap sediments, generate shade and create an architectural complexity comparable to that of terrestrial forests.

These animal forests are far more than scenic curiosities. Scientists regard them as biodiversity hotspots, nursery grounds and refugia for a vast array of associated species, from crustaceans to fish, and as key regulators of biogeochemical cycling in the sediments beneath their canopies. Yet because mesophotic depths require expensive equipment to explore, black coral forests remain among the least studied and least protected ecosystems on the planet. The new research, conducted as part of the Ocean Citizen project, a marine restoration initiative funded through Horizon Europe, set out to map benthic habitats along a stretch of Tenerife’s rocky coastline between La Arena and the southern end of San Juan Beach, and in doing so captured an unexpected ecological relationship on camera.

Between June 16 and 25, 2024, the research team, which included scientists from the University of Salento in Italy and the non-profit organization Oceana, deployed a remotely operated vehicle to survey the seabed. The vehicle, a Sibiu Pro model from Nido Robotics rated to 300 meters, was equipped with 4K cameras, LED lights and eight thrusters. It crawled along the bottom at an average speed of just 0.2 to 0.3 knots, its camera field of view spanning roughly 185 centimeters. Of 61 planned dives, 55 were completed, covering a total surveyed area of 31,217 square meters at depths ranging from 7 to 157 meters. The resulting footage was cut into 1,502 one-minute sample units, yielding 26 hours of analyzable video from which the team identified 169 distinct taxa.

The statistical analysis revealed a striking pattern. The bearded fireworm co-occurred with whip coral forests in 40 percent of the transects, appearing in the algae at the base of the black corals or in their immediate vicinity. Peak abundances of both the fireworm and the two whip coral species studied, Stichopathes gracilis and Stichopathes setacea, fell between 80 and 120 meters depth. Pearson correlation analysis confirmed that the association was not coincidental: the relationship between fireworm occurrence and S. gracilis produced a correlation coefficient of 0.895 with a p-value of 0.0027, a strong and statistically significant positive association, while the relationship with S. setacea yielded a coefficient of 0.715 with a p-value of 0.0464, also significant. All three species showed unimodal distributions across depth, peaking in the 80 to 100 meter range, with polynomial trendlines suggesting a partial overlap of their ecological niches.

Understanding why a fireworm would seek out a black coral forest requires appreciating the peculiar biology of both partners. Hermodice carunculata is a large amphinomid polychaete armed with brittle, venom-filled bristles that deliver painful stings to anything that handles it. It is an omnivorous, opportunistic feeder and scavenger, consuming decaying organic matter, and preying on corals, sea urchins, sponges and ascidians. It can also act as a reservoir for coral pathogens. Crucially, it is thermophilic and highly tolerant of environmental variation, coping with swings in temperature, salinity and oxygen. As ocean waters warm, the species has been expanding northward through the Atlantic and the Mediterranean, attacking fish caught in nets, degrading catch quality and injuring fishers, swimmers and divers. In shallow water it favors structurally complex habitats, hiding in rock cavities and algal canopies to evade predators and reduce exposure to current and light.

The whip corals, meanwhile, are ecosystem engineers of a very particular kind. Antipatharians require strong currents to feed, extending their polyp-covered stems into the flow to capture passing particles. Those same currents that sustain the forest also create a hostile hydrodynamic environment for a soft-bodied worm. The study’s authors propose a dual function for the forest: the whip corals are a product of strong currents, yet their dense stems and algal understory simultaneously provide shelter from the current’s sheer force. At the base of the corals, accumulated algae offer both refuge and a foraging ground rich in decaying organic matter, precisely the kind of organically enriched, microbially active habitat to which fireworms are known to be attracted. In essence, the fireworm appears to exploit a contradiction engineered by the corals themselves, benefiting from the very conditions that made the forest possible while escaping the conditions that make it formidable.

Perhaps the most consequential observation is what the researchers did not see. In all the analyzed video, fireworms were detected only at the base of the coral colonies, never climbing the stems. No feeding or predation behavior against the whip corals was recorded, and the coral populations appeared healthy, with no signs of damage or necrosis, despite their high density throughout the study area. This stands in contrast to recent reports of fireworms actively feeding on branching black coral colonies, concentrating on the colony tips. The evidence from Tenerife instead suggests a commensal or opportunistic relationship, in which the fireworm uses the forest as shelter and a feeding ground on associated algae and detritus rather than as prey, although the authors caution that the possibility of occasional coral consumption cannot be ruled out without further observation.

The finding carries implications well beyond the natural history of a single worm. The detection of H. carunculata at mesophotic depths extends its known vertical distribution, echoing a similar recent discovery in the same study area in which the green alga Caulerpa prolifera was recorded down to 80 meters, far deeper than previously acknowledged. As climate change pushes thermophilic, adaptable species into new depths and latitudes, understanding which habitats will harbor them, and whether vulnerable ecosystem engineers such as black corals will face new pressures, becomes a matter of urgency. Black corals are already listed in Appendix II of CITES, which regulates their international trade, and are used as indicator species for ecosystem vulnerability by regional fishery management organizations.

The authors argue that expanding knowledge of these deep, dim ecosystems is essential for conservation and for effective monitoring guidelines. Marine Animal Forests are among the most extensively distributed ecosystems on Earth, spanning from shallow mussel beds to deep cold-water coral habitats in both tropical and polar seas, yet the mesophotic black coral forests remain largely unmapped and unprotected. Determining the foraging advantages, potential risks and patterns of association between vagile species like the fireworm and the ecosystem engineers that structure these communities is, the study concludes, a necessary step toward safeguarding them. For now, the image of a venomous worm coiled in the algal shadows at the foot of a swaying black coral whip serves as a reminder that the ocean’s twilight zone is neither empty nor simple, and that its most surprising residents are only now coming into view.

Subject of Research: Co-occurrence of the bearded fireworm Hermodice carunculata with whip black coral forests in mesophotic ecosystems off Tenerife, Spain

Article Title: Co-occurrence of Hermodice carunculata (Polychaeta, Annelida) and Stichopathes spp. (Hexacorallia, Cnidaria) off Punta Blanca (SW Tenerife, Spain)

Article References: Fraissinet, S., Aguilar, R., Blanco, J., Arduini, D., Anglano, M., & Rossi, S. (2026). Co-occurrence of Hermodice carunculata (Polychaeta, Annelida) and Stichopathes spp. (Hexacorallia, Cnidaria) off Punta Blanca (SW Tenerife, Spain). Discover Conservation, 3(1), Article 23. https://doi.org/10.1007/s44353-026-00092-0

Image Credits: AI Generated

DOI: 10.1007/s44353-026-00092-0

Keywords: Hermodice carunculata, bearded fireworm, Stichopathes, black corals, mesophotic ecosystems, Marine Animal Forests, Tenerife, Canary Islands, ROV survey, biodiversity hotspots, ecosystem engineers, climate change

The work, led by Casey B. Butler of Florida International University and the Florida Fish and Wildlife Conservation Commission’s Fish and Wildlife Research Institute, together with Cheyne M. Springbett and Alastair R. Harborne, set out to test a long-standing but unverified idea. Field surveys had already shown that snail abundance tends to be lower where lobster densities are high, and protected areas with intact predator communities support fewer corallivores than fished reefs. But correlation is not mechanism: lobsters might simply suppress corallivores by eating them, or thriving lobster populations and sparse snail populations might both reflect the same high-quality reef habitat. Whether corallivores actually behave as if lobsters were dangerous — a non-consumptive, risk-induced response — had never been directly tested.

Corallivory matters because live coral tissue is a finite and increasingly scarce resource. Coralliophila galea can strip between roughly 1 and 10 square centimeters of tissue per day from key restoration species such as elkhorn coral, preferentially attacks corals that are already stressed or diseased, and can facilitate the transmission of coral diseases. The bearded fireworm is a more generalist feeder, but its grazing causes disproportionate tissue necrosis and it has been implicated as a vector of coral disease. Current management of both pests relies on divers physically removing them from restoration sites — a labor-intensive approach that works at the scale of an outplanted patch but is hopeless across an entire reef system. If predator presence alone could keep corallivores away from corals, restoration practitioners would gain a self-sustaining, ecologically grounded alternative.

To isolate the effect of chemical cues, the team ran choice experiments in custom-built clear PVC chambers, each 61 centimeters long, with lobster-conditioned seawater flowing in from one end and clean control seawater from the other. Lobsters were fed at least an hour before each trial to ensure urine production — the presumed carrier of predator odor — and then incubated in head tanks for eight hours. A central entrance port delivered the test animal into a mixing zone where the two plumes met, and the animals’ positions were recorded overnight under infrared light, since all three species are nocturnal. No coral, food, or shelter was provided inside the chambers, deliberately removing any confounding attraction and leaving only the chemical landscape of risk. Side assignments were randomized by coin flip, and preliminary trials with control water on both sides confirmed the chambers themselves carried no bias.

The results split cleanly between the two corallivores and, surprisingly, between the two lobster species. Fireworms exposed to P. argus odor showed what the authors characterize as active flight behavior: about 75 percent chose the control side, only 22 percent entered the lobster side, and video-tracking showed individuals typically made a single, rapid decision — within a median of one minute — and then held their position for the rest of the hour-long trial. Their time budgets told the same story, with a median of 98.5 percent of trial time spent in the control zone. Snails responded differently. Confronted with P. argus odor, nearly half initially selected the control side while only 7 percent moved toward the lobster cue, a statistically significant immediate avoidance. But over the course of the eleven-hour trial the initial avoidance weakened: snails increasingly congregated in the central mixing zone, which dye tests showed still carried a diluted cue, rather than holding the control side outright.

The authors interpret the snails’ behavior cautiously. For a slow-moving gastropod, staying put is not necessarily a non-response; minimizing movement is a well-documented antipredator strategy across taxa, lowering detection risk at the cost of foraging opportunity, and the pattern was consistent with sheltering in place. Yet alternative explanations remain open. The snails had been held without food for up to twelve weeks in the laboratory because they refused to feed in captivity, and prolonged starvation is known to dull chemosensory responsiveness in other gastropods. Habituation to a persistent cue, or an unmeasured drift in cue concentration as the static head tanks slowly drained over the long overnight trials, could also account for the fading response. What the data do show unambiguously is that snails spent significantly more time away from the P. argus cue than in direct contact with it.

Perhaps the most counterintuitive finding concerned the lobsters themselves. The spotted spiny lobster, Panulirus guttatus, is a reef obligate, present on the reef day and night, while P. argus migrates between reef and seagrass habitats. The team had expected the ever-present reef resident to elicit the stronger response. The opposite occurred: neither snails nor fireworms showed significant avoidance of P. guttatus cues. The authors offer several non-mutually-exclusive explanations. The two species may differ in the composition, concentration, or release frequency of the chemical cues in their urine. Predator labelling — in which a predator’s recent diet is written into its chemical signature, allowing prey to distinguish individuals that have recently eaten their own kind — could make P. argus, if it eats more molluscs and polychaetes, smell distinctly more dangerous. Alternatively, prey may have habituated to the near-constant odor of a resident species, treating it as background noise, while the episodic arrival of a migratory forager is a more reliable signal of imminent threat. Comparable patterns have been documented in mud crabs, which hide more strongly from cues of wide-ranging hunting blue crabs than from stationary ambush predators.

The broader ecological implication is that lobster presence could suppress coral tissue loss without a single act of predation. If the behavioral avoidance observed in the lab is sustained under field conditions, individual-level avoidance could scale to population- and community-level reductions in corallivory in areas of high lobster abundance. This matters against the backdrop of Caribbean-wide lobster overfishing: regional landings have fallen by roughly 20 percent, and fishing mortality in some areas has reached as high as 98 percent, raising the possibility that depleted predator populations have quietly amplified corallivory on already stressed reefs. Rebuilding lobster stocks through fisheries management and marine protected areas, the authors suggest, may help reverse corallivore-driven feedbacks that accelerate reef degradation, and restoration sites chosen for high lobster densities — or structures designed to enhance lobster habitat — could harness natural predator–prey dynamics to protect outplanted corals.

The team is careful to state the limits of the study. It measured avoidance behavior, not corallivory itself; demonstrating that lobster cues actually reduce coral tissue loss will require directly measuring feeding rates or coral survival under cue exposure, ideally in field or mesocosm settings that incorporate natural hydrodynamics and multi-predator landscapes. Cue strengths were not matched between lobster species, and lobster biomass was not a significant predictor of response in exploratory screening, but direct species-to-species comparisons of cue intensity should still be made cautiously. Identifying the specific metabolites involved, testing for context dependence, and examining how environmental stressors modulate these responses are priorities for future work. Even so, the study documents something genuinely new on coral reefs: a risk-induced pathway linking an invertebrate predator, invertebrate corallivores, and the corals caught in between. In an era when reef managers can control few of the global stressors bearing down on corals, a fear-based mechanism that could be protected, restored, or even engineered into restoration planning stands out as one of the more hopeful findings to emerge from the ecology of fear.

The study also situates itself within a broader theoretical framework known as non-consumptive predator effects, sometimes called the ecology of fear. Decades of research have shown that prey responses to perceived risk — altered foraging, shifts in habitat use, reduced movement — can propagate through food webs and indirectly benefit species at lower trophic levels. On reefs, however, most documented risk effects involve fish prey and mobile predators; evidence for chemically mediated fear responses among benthic invertebrates has remained sparse, making this experiment a notable addition.

The choice of study species reflects practical restoration concerns in the Florida Keys, where both snails and fireworms are documented pests of outplanted Acropora corals. Prior work had shown that adding the predatory rock snail Thais deltoidea to outplanting sites reduced tissue loss, but the mechanism was inferred rather than measured. The lobster study strengthens this line of inquiry by isolating chemical cues alone, demonstrating that predator odor by itself is sufficient to alter corallivore behavior.

Methodological details also matter for interpreting the findings. Trials ran overnight under infrared illumination to accommodate the nocturnal habits of all three species, and control trials with identical seawater on both sides confirmed no chamber bias. Because lobsters were incubated in static head tanks, cue concentration may have drifted over long trials — a limitation the authors acknowledge when weighing the snails’ fading response against genuine sheltering behavior.

Subject of Research: Non-consumptive effects of spiny lobster chemical cues on the avoidance behavior of Caribbean corallivorous invertebrates

Article Title: Spiny lobsters elicit avoidance behaviors in corallivorous invertebrates

Article References: Butler, C. B., Springbett, C. M., & Harborne, A. R. (2026). Spiny lobsters elicit avoidance behaviors in corallivorous invertebrates. Coral Reefs. https://doi.org/10.1007/s00338-026-02963-3

Image Credits: AI Generated

DOI: 10.1007/s00338-026-02963-3

Keywords: spiny lobster, corallivory, coral reefs, non-consumptive effects, chemical cues, predator-prey interactions, Panulirus argus, Hermodice carunculata, Coralliophila galea, coral restoration, behavioral ecology, Caribbean marine ecosystems