2.8 million deep-sea eels crowd one underwater mountain off Tasmania

Every southern autumn, more than a kilometre below the sea south of Tasmania, the flank of an extinct underwater volcano fills with eels. Sonar has now put a number on the swarm: close to three million basketwork eels, packed so tightly that their echoes merge into one solid cloud on a ship’s screen.

It is the first count of what Australian marine-park scientists describe as the only known spawning gathering of deep-sea eels anywhere. For fish whose breeding is usually inferred from scraps of evidence, a place where millions turn up on a predictable schedule gives biologists something they almost never get: an address.

The study, led by marine scientist Ben Scoulding of CSIRO, Australia’s national science agency, and published in Royal Society Open Science, estimates that the basketwork eel (Diastobranchus capensis) aggregation on Patience Seamount peaks between March and May at 3,547 to 3,557 tonnes of fish, roughly 2.8 million individuals. Divide the weight by the head count and the average eel comes out at about 1.3 kilograms.

The eels do not spread across the mountain. They crowd into dense swarms at a particular depth on its western side, over deeper and flatter ground, while orange roughy, a long-lived deep-sea fish that shares the seamount, keep to the steep slopes near the summit. Of the more than 120 seamounts in the surrounding Huon Marine Park, Patience is the only one where gatherings like this have been recorded.

Why sonar alone could not count the eels

Counting fish a kilometre down is mostly a matter of listening to echoes. The CSIRO team ran nine acoustic surveys over Patience Seamount between 2023 and 2025, including voyages on the research vessel Investigator, sending sound pulses down from the ship and measuring how strongly the water returned them. A dense school sends back a strong echo, and the stronger the echo from a given volume of water, the more fish it holds.

Sonar hears fish, not species. Orange roughy also gather at Patience, and an echo cannot say which animal produced it. So the researchers towed a deep camera system over the seamount to see which fish occupied which ground, used those images to split the acoustic signal between the two species, and collected a small number of eels to check their condition. The authors stress that “even with calibrated multi-frequency acoustics, camera imagery and physical sampling remain essential” for identifying species and estimating their size.

Orange roughy forced one more safeguard. When the team let the numbers run freely, the roughy estimate became implausible next to regional stock assessments. They therefore held roughy within a conservative range of 200 to 850 tonnes, drawn from historical assessments, before calculating the eels. The habitat split helped: eels on the deep, flat ground and roughy on the steep summit slopes were easier to tell apart on the echograms.

Why biologists think the eels come to breed

The evidence is in the fish. Most females examined carried eggs at an advanced stage, ready or close to ready for spawning, and every eel examined had an empty stomach. Animals that gather to feed have something in their guts. Many eels stop eating before they spawn, as European eels do on their long migration to the open ocean.

The calendar points the same way. The swarm grows sharply from March to May, which is autumn in the Southern Hemisphere, the pattern of a breeding season rather than a permanent residence.

Scientists have suspected this for years. Commercial trawlers that worked these seamounts before they were protected reported large catches of basketwork eels at Patience, and later research voyages caught females with well-developed eggs there even in December. That hints the eels may arrive across much of the year, spawn and leave. What nobody had was a size.

How much 3,550 tonnes of eels really is

At the peak, the eels on Patience weigh more than 20 adult blue whales, or nearly half as much as the iron structure of the Eiffel Tower. All of it gathers on one corner of a small volcanic cone whose base covers about three square kilometres, rising from about 1,250 metres deep to a summit roughly 857 metres below the surface.

The comparison that matters most to eel biologists is with the eels everyone knows. European and American eels cross the Atlantic to breed in the Sargasso Sea, a fact pieced together from tiny larvae caught there. Nobody has ever watched them spawn. It took satellite tags to follow adult European eels into the Sargasso Sea at all, a result reported in Scientific Reports. A breeding site where millions of eels gather at a known place, a known depth and a known season offers something those famous species never have.

The basketwork eel belongs to a different family, the cutthroat eels, which live their whole lives in the deep ocean. At Patience it spawns around 1,100 metres down, in water that sunlight never reaches.

Orange roughy add a conservation story. The species once supported one of Australia’s largest and most valuable fisheries, and it can live for more than a century, so heavily fished populations recover slowly. The authors write that “Patience Seamount’s protection and recovery from historical fishing provide a rare opportunity for long-term study of deep-sea ecological processes.”

What the count does not settle

The total is an estimate built on assumptions. Turning echoes into tonnes requires knowing how strongly a single eel reflects sound. Splitting the echo between eels and roughy leans on camera tows that see only part of the seamount and on a roughy range borrowed from older stock assessments. The authors found that without those limits the numbers drift into the implausible, which shows how much the answer depends on them. Different assumptions would move the figure.

The spawning case is strong but indirect. Advanced eggs and empty stomachs are what a breeding gathering should look like, yet no one has watched basketwork eels release eggs, and their larvae have not been traced back to the seamount. Where the eels come from, how long each one stays and where the fertilised eggs drift afterwards remain open questions.

Then there is the puzzle of place. Patience is one seamount among more than 120 in the park, many of them similar volcanic cones at similar depths. Why the eels choose this one, and apparently only this one, the study does not answer.

Common questions about the Patience Seamount eels

What is a basketwork eel?

The basketwork eel, Diastobranchus capensis, is a deep-sea eel of the cutthroat eel family that lives near the seafloor. At Patience Seamount, south of Tasmania, it gathers around 1,100 metres below the surface in what scientists believe is a spawning aggregation.

Where is Patience Seamount?

It is an extinct underwater volcano in the Huon Marine Park, south of Tasmania, Australia. Its base lies about 1,250 metres down and covers roughly three square kilometres, and its summit sits about 857 metres below the surface.

Why do millions of eels gather there?

Most likely to spawn. Most females examined carried advanced eggs and the eels had empty stomachs, which fits a breeding gathering rather than a feeding one. Why they choose Patience and not the many similar seamounts nearby is unknown.

Has anyone seen eels spawn in the wild?

Not the famous ones. European and American eels breed in the Sargasso Sea, but their spawning has never been directly observed. Patience Seamount is described as the only known spawning aggregation of deep-sea eels, which makes it a rare natural laboratory.

What comes next

The seamounts have been protected in stages: a no-trawl zone in 1995, the Tasmanian Seamounts Marine Reserve in 1999, and the larger Huon Marine Park when Australia’s South-east Marine Parks Network was declared in 2007. That same year, CSIRO multibeam sonar mapped 123 seamounts in the area and its scientists documented the eel gathering. The new estimate draws on nine acoustic surveys between 2023 and 2025, with camera frames from an April 2025 voyage, and the paper appeared on 28 October 2026.

The authors call Patience Seamount “a rare and ecologically significant site, offering a unique opportunity for long-term monitoring,” and present their approach as “a scalable, non-invasive framework for monitoring deep-sea aggregations in marine protected areas.” Each future survey can now test one thing a single count cannot: whether close to three million eels is a ceiling, or a population still recovering from the trawl nets.

Reference: Scoulding et al., “A unique ecological phenomenon quantified by hydroacoustics: the deep-sea eel spawning aggregation at Patience Seamount,” Royal Society Open Science, 2026. DOI: 10.1098/rsos.260397

Tags: CSIRO, deep sea, eels, Huon Marine Park, marine biology, Patience Seamount