Industries / Food, farming and forestry / Fishing and aquaculture / Shellfish farming
Disease
This is the work of finding, tracking and limiting the infections that kill farmed shellfish, often in mass die-offs.
7
Milestones achieved, from 1950 to 2016.
3
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1950
Dermo parasite named
Mackin, Owen and Collier described a single-celled parasite in eastern oysters from the US Gulf of Mexico coast; it is now called Perkinsus marinus, and the disease it causes, Dermo. In Chesapeake Bay it had been recognised in 1949 as a cause of late-summer oyster deaths, and around 1990 it spread north to Delaware Bay and New England, by Fisheries and Oceans Canada's account.
LessonFisheries and Oceans Canada links the northward spread to oysters being moved, combined with warmer winters.
Credit
Mackin, Owen and Collier (Science, 1950)
1957
MSX and resistant breeding
In 1957 an introduced parasite, Haplosporidium nelsoni, began killing oysters in Delaware Bay; the disease became known as MSX. In the late 1950s and early 1960s, deaths of adult oysters approached 100% over three years in the saltier parts of Delaware and Chesapeake bays, according to Fisheries and Oceans Canada. Harold Haskin at Rutgers began breeding oysters that survived, producing strains highly resistant to MSX.
Lessonresistant oysters can still carry the parasite, so breeding does not remove the need to control where oysters are moved.
Credit
Harold Haskin and Susan Ford, Rutgers University (Haskin Shellfish Research Laboratory)
1971
France replaces its oyster
From 1967 a gill disease, later traced to a virus, struck the Portuguese oyster that French farms grew, and between 1970 and 1973 the species nearly vanished from the French coast. Losses reached 60,000 tonnes of adult oysters a year and nearly 5,000 growers were affected. From 1971 to 1975 France imported Pacific oyster breeding stock from British Columbia, in "Operation Résur", and spat from Japan, and Pacific oyster output reached 80,000 tonnes by 1976.
Lessoneach shipment was dipped in fresh water to kill pests riding along, which the authors credit with limiting the species introduced with the oysters.
Credit
French oyster growers and government scientists; account by Grizel and Héral (Ifremer)
1979
Bonamia strikes flat oysters
French scientists described a new parasite living inside the blood cells of the European flat oyster and named it Bonamia ostreae. With an earlier parasite, Marteilia refringens, it cut French flat oyster output from about 20,000 tonnes a year in the 1970s to 1,800 tonnes in 1995, by Fisheries and Oceans Canada's figures, which say it most likely reached Europe from California in the late 1970s with moved oysters. The World Organisation for Animal Health (WOAH) keeps a diagnostic standard for it.
Lessonits spread within Europe has been linked above all to moving live shellfish.
Credit
Y. Pichot, M. Comps, G. Tigé, H. Grizel and M. A. Rabouin, who described the parasite
2008
Herpesvirus kills young oysters
In summer 2008 French farms reported deaths of 80 to 100% among Pacific oysters. Ifremer scientists linked them to a new form of ostreid herpesvirus 1, which they named OsHV-1 µVar. In March 2010 the European Commission required testing when unusual deaths occur, containment areas around them, and health certificates for Pacific oysters leaving those areas.
Lessonthe European Food Safety Authority judged the virus a necessary but not sufficient cause, with sudden warming and moving untested spat among the risk factors, and advised checking spat health at the source.
Credit
Ifremer (Segarra and colleagues); European Commission Regulation (EU) No 175/2010
2010
POMS reaches Australia
Between November 2010 and January 2011, farmed Pacific oysters in the Georges River, New South Wales, suffered more than 95% mortality, and tests found a variant of the same herpesvirus. The disease, Pacific Oyster Mortality Syndrome (POMS), reached southern Tasmania in January 2016 and killed up to 95% of oysters on infected farms; Tasmanian hatcheries supplied about 90% of South Australia's oyster spat, so the outbreak also cut seed supply.
LessonSydney rock oysters growing beside the dying Pacific oysters were not affected.
Credit
Jenkins and colleagues (2013), who identified the virus in the Georges River oysters
2016
Bred resistance restores Tasmania
Australian Seafood Industries (ASI) had bred Pacific oysters since the late 1990s and began selecting for POMS resistance after the virus reached New South Wales. After the 2016 Tasmanian outbreak, oyster families were exposed to the disease on farms and the survivors bred the next generation; by the end of summer 2019 Tasmania's Pacific oyster stocks had almost fully recovered, by ASI's account. In March 2021 South Australia approved bringing in ASI's most resistant families after virus screening and a month of quarantine.
Lessonsome resistant survivors still carry the virus, which is why spat were screened and quarantined before crossing a state border.
Credit
Australian Seafood Industries (Matt Cunningham); Institute for Marine and Antarctic Studies, University of Tasmania
Watch for
A vaccine for oysters
University of Sydney researchers report that in a laboratory challenge with the POMS virus, 92% of oysters given their best candidate vaccine survived, against 43% of unvaccinated oysters. They are filing a patent and seeking funding to work out how to deliver it and how long protection lasts.
Achieved whenfarm trials, reported by an outside party, show vaccinated oysters surviving natural POMS outbreaks better than unvaccinated ones, with a delivery method that works at hatchery scale.
Claimed by
Francisca Samsing and Arman Hossain, University of Sydney, at the NSW Oyster Conference, May 2026
A new virus suspect
Pacific oysters on the Hawkesbury River, New South Wales, died within two to three weeks at each site in summer 2025–26, yet tests found no POMS that summer or the one before. University of Sydney researchers point to a possible role for PoSV-1, a previously unknown virus found in 2024 by sequencing all the RNA in samples, alongside water above 25 °C, Vibrio bacteria and long-lasting inflammation.
Achieved wheninfection trials reproduce the disease with PoSV-1 and an independent laboratory confirms the link.
Claimed by
University of Sydney (Francisca Samsing, Richard Whittington), May 2026
Resistant seed for Atlantic Canada
Fisheries and Oceans Canada confirmed Dermo in Canada for the first time in November 2024, in New Brunswick and Nova Scotia, and in Newfoundland in June 2025. In October 2026 Rutgers licensed its MSX- and Dermo-resistant breeding stock to Atlantic Aqua Farms, which says it has produced more than 100 million seed from Rutgers lines for Prince Edward Island and New Brunswick and that the seed has "performed extremely well so far".
Achieved whensurvival data from Canadian farms over at least one full disease season, gathered by an outside party such as Fisheries and Oceans Canada, show lower losses than in local oysters.
Claimed by
Rutgers University (Ximing Guo) and Atlantic Aqua Farms, October 2026
Open
01
Score dermo slides by camera
A camera-and-software tool that reads Ray's fluid thioglycollate medium (RFTM) slides, in which a piece of oyster tissue is cultured for about a week and stained with iodine so the Dermo parasite shows as dark round cells, and gives the 0 to 5 Mackin score a trained reader would.
The briefCredit · unclaimed02
Count dead oysters in bags
A tool that counts dead oysters (gaping shells, or empty shells still joined at the hinge) in photos of bags, baskets or trays taken during routine farm checks, so a farm keeps a mortality record for every unit and spots a die-off early.
The briefCredit · unclaimed
Open 01 · the brief
Score dermo slides by camera
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A camera-and-software tool that reads Ray's fluid thioglycollate medium (RFTM) slides, in which a piece of oyster tissue is cultured for about a week and stained with iodine so the Dermo parasite shows as dark round cells, and gives the 0 to 5 Mackin score a trained reader would.
Why it is still open
Readers still score these slides by eye, scanning under the microscope and comparing with reference photographs, as set out in the guide used by the Oyster Sentinel monitoring network. Two searches found no published image tool for scoring them.
Built with what exists
A microscope with a phone or USB camera; slide photos with scores from a lab that already runs RFTM tests; an image classifier trained in PyTorch; a quantitative PCR test (one that counts parasite DNA) on part of each sample as a second check.
First step
Ask a lab that runs RFTM tests to photograph each slide it reads for one month, alongside its score.
Done looks like
1,000 scored slides from at least two labs; on 300 kept out of training, the tool's score is within one point of the expert's for at least 90%, and it never scores a heavy infection (3 or more) as light (1 or less).
Open 02 · the brief
Count dead oysters in bags
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A tool that counts dead oysters (gaping shells, or empty shells still joined at the hinge) in photos of bags, baskets or trays taken during routine farm checks, so a farm keeps a mortality record for every unit and spots a die-off early.
Why it is still open
In a June 2025 test on oyster reef images, an AI model identified live oysters less accurately (63%) than human experts (74%) and non-experts (75%), and the authors wrote that "future research should address methods that improve the detection of living vs dead oysters".
Built with what exists
Phone or action cameras; CVAT for labelling; Meta's Segment Anything 2 to outline shells; an object detector such as Ultralytics YOLO.
First step
At one farm, photograph 30 bags before the crew empties and counts them, and record the hand count of dead oysters for each.
Done looks like
2,000 labelled oysters photographed in gear at three farms; on bags kept out of training, the tool's dead count is within 10% of a hand count for at least 90% of bags; images and labels published.