BRENNEN

Industries / Food, farming and forestry / Fishing and aquaculture / Shellfish farming

Cleaning and safety testing

This is the work of making shellfish safe to eat after harvest: holding them in clean water so they purge germs, and testing their flesh for toxins, bacteria and viruses.

7

Milestones achieved, from c. 1914 to 2023.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. c. 1914

    Purifying mussels at Conwy

    From about 1914 a station at Conwy, North Wales, purified mussels by a method R. W. Dodgson worked out there: the mussels sit on wooden grids and get two soaks of about 16 to 20 hours each in seawater sterilised with chlorine and then dechlorinated, so they pump their own guts clean. In 1950 this was still the process at the British government's mussel-cleansing stations. Holding shellfish in clean water this way is now called depuration.

    Lessonchlorine left in the water slows the mussels' pumping, so 1950 trials kept it at or below 0.05 parts per million before the mussels went in.

    cambridge.orgcoflein.gov.uk

    Credit

    R. W. Dodgson, for the Ministry of Agriculture and Fisheries

  2. 1937

    Mouse test for paralytic poison

    H. Sommer and K. F. Meyer published a bioassay, a test on live animals, for paralytic shellfish poisoning (PSP): mice are injected with an acid extract of shellfish, and the time until they die is converted through a table into units of toxin. AOAC International, a standards body for laboratory methods, adopted it in 1959. It was the regulatory standard for decades, and a 2025 review found that most US laboratories still used it.

    Lessonits strength is that it responds to any PSP toxin, even one whose chemical structure is unknown.

    pmc.ncbi.nlm.nih.gov

    Credit

    H. Sommer and K. F. Meyer; AOAC International

  3. 1978

    Norovirus survives purification

    At least 2,000 people across Australia fell ill after eating oysters in June and July 1978. The cause was Norwalk virus, now classed as a norovirus, which the investigators said had not been linked to food-borne illness before. Purification was then introduced, yet between December 1978 and August 1979, 52 volunteers who test-ate purified oysters from the Georges River, New South Wales, fell ill, mostly in two weeks after heavy rain.

    Lessonpurification that removes gut bacteria does not reliably remove viruses; FAO's 2008 depuration manual still calls it "less effective" against norovirus and hepatitis A.

    ebi.ac.ukebi.ac.ukfao.org

    Credit

    A. M. Murphy, G. S. Grohmann and colleagues

  4. 1987

    Amnesic poisoning from mussels

    In late 1987, 107 people met the case definition for a new illness after eating farmed mussels from three estuaries in eastern Prince Edward Island, Canada. Nineteen were taken to hospital, a quarter lost short-term memory, and three died, by one study's count; a second study in the same journal issue examined four who died within four months. Chemists traced the illness to domoic acid, a toxin that overstimulates nerve cells, made by a diatom (a single-celled alga). A 2000 review calls this the first recorded case of amnesic shellfish poisoning, and Health Canada now sets a limit of 20 mg of domoic acid per kg of shellfish.

    Lessonnone of the known shellfish toxins were involved, so a new toxin had to be identified and added to routine testing.

    ebi.ac.ukebi.ac.ukpmc.ncbi.nlm.nih.govcanada.ca

    Credit

    J. L. C. Wright and colleagues (identified the toxin); T. M. Perl and colleagues (outbreak study, 1990)

  5. 2011

    Chemistry replaces the mouse test

    Commission Regulation (EU) No 15/2011 made liquid chromatography with tandem mass spectrometry (LC-MS/MS), a laboratory method that separates toxin molecules and identifies each by its mass, the EU reference method for fat-soluble shellfish toxins, and allowed mouse tests for them only until 31 December 2014. Paralytic toxins followed: the EU accepted a chemical method by Lawrence and colleagues (AOAC 2005.06) as an alternative in 2006, and in 2021 said the mouse bioassay for these toxins should be discontinued.

    eur-lex.europa.eueur-lex.europa.eupmc.ncbi.nlm.nih.gov

    Credit

    European Commission; EU Reference Laboratory for marine biotoxins; Lawrence and colleagues

  6. 2017

    A standard norovirus test

    ISO 15216-1:2017 set an international laboratory method for measuring norovirus and hepatitis A virus in foods, including bivalve shellfish, using real-time RT-PCR, which copies and counts the viruses' genetic material. It replaced a 2013 technical specification.

    Lessona shared test came before any legal limit; Irish guidance revised in 2025 still notes that shellfish law sets no limit for norovirus.

    iso.orgfsai.ie

    Credit

    ISO technical committee ISO/TC 34/SC 9

  7. 2023

    Farm toxin test withdrawn

    Neogen made a lateral-flow test, a strip read like a home COVID test, that let growers screen oysters for paralytic toxins in under 20 minutes. Tasmanian farms used it to check batches at harvest, and South Australia used it to screen low-risk product for regulators. In 2023 Neogen stopped making it, catching Australian, UK and New Zealand shellfish industries "off-guard", and University of Tasmania trials published in 2025 found that neither of two replacements yet met all five performance criteria.

    Lessona safety check that depends on one company's product can vanish when that company stops making it.

    oystersaustralia.org

    Credit

    Andreas Seger and Alison Turnbull, Institute for Marine and Antarctic Studies, University of Tasmania

Watch for

  1. Pulling norovirus from tank water

    Coldep, a French water-treatment company, says its VAL AQUA system cut norovirus in seawater by 95% in 120 minutes, from 37 to 8 million genome copies per litre, by lifting viruses out on tiny bubbles. It is meant for shellfish purification tanks, and the trials ran in 2025 and 2026 at a marine station in Sète, France, under a protocol reviewed by Ifremer, France's marine research institute. The claim concerns the water, not the oysters.

    Achieved whenan independent or peer-reviewed trial shows less norovirus in oysters purified with the system than without it, and a food-safety authority accepts it in approved purification centres.

    aquaculturemag.com

    Claimed by

    Coldep, April 2026

Open

  1. 01

    Show the bottom layer filters

    Small valve-gape sensors (a magnet glued to one shell and a magnetic-field sensor to the other, which together measure how far the shell is open) fitted to shellfish at the top, middle and bottom of a loaded depuration tank or bin, logging through a whole purification cycle, alongside E. coli tests on each layer.

    The brief
    Credit · unclaimed
  2. 02

    Test every toxin strip alike

    A public comparison of every on-farm test strip for paralytic shellfish toxins now on sale, run blind on the same split samples, which cover the main toxin mixtures found in different regions, and scored against the five criteria the University of Tasmania team set out.

    The brief
    Credit · unclaimed

Open 01 · the brief

Show the bottom layer filters

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

Small valve-gape sensors (a magnet glued to one shell and a magnetic-field sensor to the other, which together measure how far the shell is open) fitted to shellfish at the top, middle and bottom of a loaded depuration tank or bin, logging through a whole purification cycle, alongside E. coli tests on each layer.

Why it is still open

Depuration works only while shellfish are open and pumping. FAO's 2008 manual says that in deeper systems used for mussels "there is no direct evidence that the individuals at the very bottom are able to open", and that a UK bulk-bin system "has not been verified for other species". A search found gape sensors used to study shellfish in the wild and in laboratories, but none used to check depuration loads.

fao.org

Built with what exists

Hall-effect sensors and small neodymium magnets; an Arduino or Raspberry Pi Pico logger in a waterproof box; an accredited food laboratory for E. coli counts.

  1. First step

    Build four sensors, fit them to oysters in a home aquarium, and check that the record shows them opening and closing.

  2. Done looks like

    Ten full cycles at a working depuration plant, with gape records from at least 24 animals per cycle across three layers and E. coli results for each layer before and after; the share of time each layer spends open, and its E. coli drop, published.

Open 02 · the brief

Test every toxin strip alike

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public comparison of every on-farm test strip for paralytic shellfish toxins now on sale, run blind on the same split samples, which cover the main toxin mixtures found in different regions, and scored against the five criteria the University of Tasmania team set out.

Why it is still open

Since Neogen stopped making its strip, makers in Chile, Canada and the US have offered replacements, but the University of Tasmania team concluded in April 2025 that "additional pilot trials are required to select a suitable replacement", tested only two kits, and noted that toxin-positive oyster tissue for testing is scarce. A strip's antibodies detect some toxin forms better than others, so results depend on the region's toxin mixture.

oystersaustralia.org

Built with what exists

The kits now on sale (the report names AquaBC, Sensoreal and Attogene); split samples and reference results by LC-MS/MS from a partner biotoxin laboratory, which handles all toxic material; a published blind-testing protocol.

  1. First step

    Ask a biotoxin monitoring laboratory whether it can share frozen split samples, and write the blind-testing protocol with it.

  2. Done looks like

    At least four kits, each tested on samples covering at least six toxin mixtures at 0, 0.2, 0.4, 0.6 and 0.8 mg STX-equivalents per kg with five replicates each, plus 30 toxin-free samples; a public table showing which of the five criteria each kit meets.