Almost 4,000 kilometres separate Shark Bay on Australia’s west coast from Hervey Bay on its east. Bottlenose dolphins in both places have nevertheless arrived at the same unusual way to catch a fish: use a large empty sea-snail shell as a trap, carry it to the surface and drain the prey toward the mouth.
The behaviour is known as shelling. Researchers have followed it in Western Australia’s Shark Bay for years. Now a new open-access study in Marine Mammal Science presents the first published records from Queensland’s Great Sandy Marine Park.
The huge geographic separation supports an arresting interpretation: dolphins on the two sides of Australia appear to have invented the same tactic independently. But the wording matters. The researchers documented matching behaviour in distant populations; they did not run a test that could reconstruct the exact first moment of invention on either coast.
Shelling turns a fish’s refuge into a trap
A small fish being pursued near the seabed can disappear into the opening of a dead gastropod shell. For most predators, that refuge would end the chase. A shelling dolphin instead inserts its rostrum into the aperture, raises the heavy shell and tips it above the water. Water drains out, bringing the trapped prey within reach.
The first detailed scientific account of dolphins carrying conch shells came from Shark Bay. Photographic sequences showed dolphins lifting large shells, with a fish visible as the shell was emptied. Later observations established that the tactic is rare, fast and easy to overlook.
It also requires more than simply carrying an object. A dolphin must locate a suitable dead shell, exploit a fish’s tendency to hide inside it, keep the aperture oriented during the ascent and drain it without losing the meal. Researchers consequently describe the shell as a tool and the tactic as a specialised form of foraging.
Three Queensland dolphins supplied the clearest evidence
Alexis Levengood and colleagues have conducted opportunistic boat surveys in the Great Sandy Marine Park since 2022. Their paper centres on two encounters in Hervey Bay during 2025, both involving Indo-Pacific bottlenose dolphins, locally known as buthu in the Butchulla language.
On 6 July, the team approached an adult female named Stevie Nicks. She used her rostrum to position a baler shell, Melo amphora, at the surface and lifted the entire shell out of the water across three surfacings. Rain ended the follow after nine minutes. The observers did not see a fish leave the shell and could not determine whether the attempt produced a meal.
On 1 October, a mother named Maserati surfaced repeatedly with another baler shell while a drone filmed from above. The video captured a fish escaping, followed by Maserati dropping the shell and chasing it. Whether she caught the fish remained hidden beneath poor-visibility water.
Three minutes later, Maserati’s calf Bentley surfaced with the same shell in the same place. The calf put its rostrum into the opening, raised the shell and slapped it against the surface. A photograph appears to show the body or tail of a suspected lizardfish protruding from the aperture. Again, the researchers could not confirm that either dolphin ate the prey.
The paper also includes photographs from an ecotour vessel taken in 2013 and 2019 in the same general area. Those records lack precise GPS positions and behavioural detail, but they suggest that Queensland shelling was present years before the 2025 research encounters.
Why opposite coasts suggest two inventions
Before the new paper, all published shelling observations came from Shark Bay. Hervey Bay lies roughly 4,000 kilometres away in a different dolphin population on the opposite side of the continent. There is no plausible chain of ordinary dolphin associations connecting the two study sites.
That is why the authors say their observations contribute evidence of independent innovation across Australia’s west and east coasts. Similar ecological ingredients may have offered similar opportunities: large empty shells on the seabed, small fish that use those cavities as shelter, and dolphins capable of manipulating awkward objects.
The conclusion is persuasive but inferential. The study did not compare genetic lineages and social networks across the continent, nor can it rule out unrecorded occurrences elsewhere. “Independently learned” means the Queensland population could not reasonably have copied Shark Bay dolphins directly. It does not mean every dolphin within each population reinvented the tactic alone.
That distinction fits broader work on aquatic tools. A review of tool use by aquatic animals notes that water changes both the physical demands and the opportunities for object manipulation. Shelling is a compact example: buoyancy helps move a large shell, while gravity becomes useful only when the dolphin lifts it clear of the surface.
At Shark Bay, the tactic travels between associates
An invention can happen independently and still become cultural once it appears. A 2020 study of Shark Bay’s shelling dolphins analysed 5,278 group encounters recorded from 2007 to 2018. Researchers documented 42 shelling events performed by 19 individuals.
Network-based diffusion analysis found the strongest support for transmission through social associations. Dolphins were estimated to be 10 to more than 300 times more likely to acquire shelling through their social network than through genetic inheritance or learning alone. This was the first quantitative evidence of a toothed-whale foraging tactic spreading outside the mother-calf bond.
The result contrasts with another Shark Bay tradition. Some dolphins carry marine sponges over their rostra while probing rubble for buried fish, a technique for which genetic and behavioural evidence supports maternal cultural transmission. Shelling appears more flexible, capable of moving horizontally between peers or obliquely from older non-parents to younger animals.
Even the 2020 analysis left room for separate discovery. Its best model estimated that about 43 per cent of dolphins may have acquired shelling outside the measured social network. The authors cautioned that missed encounters could make asocial learning appear more common than it was.
Bentley may have learned by watching Maserati
The Hervey Bay mother-calf sequence adds a possible transmission route. Bentley picked up the same shell three minutes after watching Maserati use it, placed the rostrum in the same opening and attempted similar movements. The new paper calls this potential evidence of vertical social learning, meaning information passed from parent to offspring.
It is not conclusive evidence of deliberate teaching. Teaching normally requires evidence that a knowledgeable animal changes its behaviour in the presence of a learner, incurs some cost or receives no immediate benefit, and helps the learner acquire a skill faster. The Queensland researchers observed one mother-calf sequence and did not test those conditions.
The calf may have learned by observation, been drawn to an object its mother had just handled, or already had experience with shells. The researchers also could not confirm a successful catch. Their cautious conclusion is that the event is consistent with maternal social transmission and deserves further observations.
That restraint is important when discussing intelligent animals. SpaceDaily has previously examined how bottlenose dolphins develop and copy signature whistles. Sophisticated communication and long maternal care make social learning plausible, but they do not let observers infer a specific mental process from one striking scene.
Invention and culture can be the same continuing story
The most interesting interpretation is not that every shelling dolphin solved the puzzle alone. It is that a useful behaviour may originate more than once, then travel locally through different social routes.
Tool use begins with an innovation. It becomes a tradition when other animals acquire and retain it partly through social learning. Shark Bay supplies quantitative evidence that shelling spreads among associates. Hervey Bay supplies a geographically separate occurrence, older photographic records and one close sequence in which a calf copied its mother’s handling of a shell.
The evidence from Queensland remains small: two detailed 2025 encounters, three dolphins and two earlier photographs. Shelling is so brief and rare that more practitioners may remain unseen. Longer observation, individual life histories and social-network data will be needed to learn how often it occurs and how calves or peers acquire it.
For now, the two coasts present a compelling natural parallel. Similar prey, similar shells and similarly flexible minds appear to have produced one remarkably similar invention. Once discovered, the trick may cease to belong to one clever dolphin and begin moving through a population as culture.