Aquaculture has won the argument about scale. In 2024 farms produced 103 million tonnes of aquatic animals, passing 100 million for the first time and overtaking wild capture, which has been stuck between 86 and 94 million tonnes since the late 1980s. FAO expects aquatic animal production to reach 214 million tonnes by 2034, with farms supplying 119 million.

What it has not won is the argument about how. From London Underground adverts to Canadian court challenges, the fights now are over fish welfare, wild fish, the wider environment and who gets to decide. China alone grows 56% of the world’s farmed aquatic animals, yet the sharpest public scrutiny falls on a far smaller sector: farmed salmon.

Dying in the pens

Mortality has become the number the public understands, and no two countries report it the same way.

Norway’s trend is improving: sea-phase mortality fell from 16.7% in 2023 to 14.2% in 2025, and the government’s stated goal is 5%. But one toxic algal bloom in March 2026 pushed southern Norway’s half-year rate to 15.8%.

Scotland shows how one industry can produce opposite headlines. Campaigners cite the official survey; producers cite record monthly survival. Both are real numbers, measured differently, and without one audited definition each side picks the one that suits it.

Chile adds a health dimension. More than 96% of its antibiotic use targets a single bacterial disease, SRS, while Norway reports using virtually none.

Escapes, lice and wild salmon

Wherever farmed and wild Atlantic salmon share a coastline, escapes and lice are the core charge, and the escapes keep happening. In February 2025 a pen sank in northern Norway and 27,000 salmon got out. Fewer than 5% were recaptured, and the escapees later raised the share of farmed fish in rivers in Troms. In Iceland about 3,500 fish escaped from a single farm in 2023 and turned up in rivers across the northwest.

The official figures probably understate the problem. One study estimated Norway’s real escapes in 2005–2011 at about one million fish a year, more than double the 413,000 that were reported.

Where failures pile up, bans follow. Washington State’s 2017 net-pen collapse released about 250,000 Atlantic salmon into Puget Sound. By 2025 the state had permanently banned commercial net pens, joining California, Oregon and Alaska.

Lice remain the most contested science. In Norway farmed fish make up 99.6% of salmon-louse hosts, and release trials in Ireland and Norway attributed on average 39% of marine mortality to lice. In British Columbia, though, removing farms has not produced a clean drop in lice on wild juveniles, a dispute examined in No Farms, Still Lice?.

Beyond the pen

Some of the hardest fights are about what farming does to everything around it. In Tasmania, Macquarie Harbour is the only home of the Maugean skate, with an estimated 40 to 120 adults left. Federal conservation advice named salmon farming as the main human contributor to the harbour’s oxygen loss, and officials advised revoking the 2012 expansion approval. In 2025 the government instead let farming continue, backed by AUD 18.3 million for artificial oxygenation. ASC and the RSPCA no longer certify salmon from the harbour.

Climate is turning algal blooms and marine heatwaves into routine business risks. In March 2026 a bloom killed nearly 1.7 million salmon in one Norwegian production area. Chile is bracing for El Niño: in 2016 a bloom there killed more than 40,000 tonnes of fish, much of it dumped at sea, and officials have now told parliament they may bury mass mortalities in land trenches next time.

Then there is feed. IFFO forecasts 2026 fishmeal output below 4.6 million tonnes and fish oil at 1.15 million tonnes, both the lowest since 2016, after Peru’s anchovy fishery faltered. Every growth plan for carnivorous species depends on replacing marine ingredients at scale.

Who decides

Increasingly, the licence to operate is set by politics rather than biology. Canada has ordered open-net salmon farming out of British Columbia’s coastal waters by 30 June 2029. Producers contest the lice science behind the decision, and some First Nations that partner with farms, such as the Kitasoo Xai’xais, are demanding a reversal. South of the border, Washington’s ban survived its last legal challenge in 2026, when the industry dropped its lawsuit.

Norway shows a different lever: tax. It introduced a resource rent tax on salmon farming in 2023, and in June 2026 parliament voted 54 to 47 to scrap the benchmark-price system used to calculate it from 2027, while keeping the tax itself. Iceland, meanwhile, missed its spring 2026 target for a new aquaculture law. The bill stalled in committee under fire from both industry and anglers, and one major producer paused expansion, citing the uncertainty.

The result is a ceiling on supply. Mowi now expects long-term global salmon supply growth of only 1–2% a year, citing regulatory limits and technology constraints. For salmon, regulation, not demand, sets the limit.

Land-based and offshore: promise meets the balance sheet

Closed systems are pitched as the answer to escapes, lice and local pollution. The biology is improving, but the economics remain hard. Atlantic Sapphire, the Florida pioneer, lost USD 30.9 million in the first half of 2026 and is being taken private at NOK 0.80 a share. Kontali says land-based farms need consistent harvest weights above 4 kg to close the profitability gap with sea cages.

Public money is filling part of the gap. Canada has put CAD 14.5 million into floating closed-cage technology in British Columbia, and Scotland granted Mowi £1.78 million for a semi-closed trial. Closed and offshore farming will add capacity, but neither is close to replacing the open net pen at scale.

Trust is not a marketing problem

The common thread is credibility. In October 2026 a campaign against Marks & Spencer put more than 5,000 adverts on the London Underground, demanding that it stop selling farmed salmon. In Chile, NGO figures cited by the Guardian count 83 worker deaths in salmon farming from 2013 to 2025, against three in Norway over 34 years.

The industry answers with its own numbers, and its critics with theirs. Better messaging will not end that stalemate. What will is the same audited figures published everywhere: mortality by cohort, lice abundance on farms and on wild fish, escapes, antibiotic use and worker injuries.

Aquaculture will keep growing because the world needs the protein. Whether farmed salmon keeps growing in open water, and on whose terms, depends on whether those numbers can be trusted.