Toroī with pūhā or watercress

What is toroī?

Toroī is a traditional Māori food combining shellfish, particularly mussels, with edible greens such as pūhā. Some documented versions use watercress instead.

The spelling toroī, with a macron over the final vowel, is recognised in Māori language references. Scientific publications also use toroi or tiroi.

In te reo Māori, the verb toroī can mean to ferment or brew. The noun also identifies a food made with mussels and pūhā.

The mussel most relevant to modern research is the New Zealand green-lipped mussel, Perna canaliculus.

Pūhā refers to edible sow-thistle species, including Sonchus oleraceus and Sonchus asper in relevant botanical accounts.

Watercress is commonly identified as Nasturtium officinale.

The food may be served as a relish or soup. Some historical methods involve storage and fermentation, but not every documented toroī preparation follows the same microbiological process.

That distinction is especially important because toroī has been the subject of serious food-safety research.

The recipe in this record reproduces a specific 2007 research preparation. It is not a validated household fermentation method.

Toroī in Māori food culture

Māori food traditions include a wide range of marine, freshwater, forest and cultivated resources.

Shellfish are particularly important in coastal and estuarine food systems.

Mussels, pāua, pipi, cockles and other marine foods have long been gathered and eaten.

Edible wild greens provide another important category of food.

Pūhā is frequently associated with Māori cooking and with dishes combining greens and meat or other proteins.

Toroī brings these two resources together.

Te Ara, the Encyclopaedia of New Zealand, describes toroī as fresh mussels combined with pūhā and mussel stock to produce a distinctive relish or soup.

This culinary description is broader than a single laboratory fermentation recipe.

It illustrates that toroī has a place in the food culture independent of the modern scientific effort to understand its microbiology.

Historical evidence and language

The precise invention date of toroī is unknown.

The Māori dictionary records the word toroī as a verb meaning to ferment or brew and as a noun referring to fermented food.

One cited written example of the verb dates to 1864.

A separate biographical text from 1998 includes toroī among recognised Māori foods.

These references demonstrate historical terminology and continuing cultural recognition.

They do not establish that the exact mussel-and-watercress formula used in a 2007 university thesis existed in the same form in 1864.

The distinction between a word's history and a particular recipe's history is important.

Traditional Māori food knowledge also contains numerous local practices, and some techniques were transmitted orally.

The absence of an early published recipe does not prove that a food is recent. It means the surviving documentary evidence should not be extended beyond what it supports.

Traditional processing and variation

Research published in 2001 examined traditional Māori food-processing methods, including toroī or tiroī.

The investigators consulted knowledgeable Māori elders and converted their accounts into process diagrams.

They identified three distinct methods of tiroī preparation.

The methods differed in how the greens and mussels were cooked, mixed and stored.

Some involved substantial cooking.

Others involved storage arrangements that might permit fermentation.

In one preparation, cooked greens were combined with mussels and sealed beneath a layer of fat.

The purpose of that layer was not established with certainty.

It could have reduced contamination from outside, but it also could have altered oxygen availability.

The researchers concluded that the available descriptions were insufficient to establish the microbiology of every process.

This is an essential historical finding: toroī cannot accurately be described as always undergoing one identical lactic-acid fermentation.

Why the fermentation claim needs care

Many preserved foods undergo changes during storage.

Some changes are caused by microorganisms.

Others are enzymatic or chemical.

A food can develop a stronger flavour over time without necessarily undergoing a protective lactic-acid fermentation.

The 2001 microbiological research found evidence that one investigated tiroī method may involve lactic-acid fermentation.

The other two appeared closer to cooking and bottling processes.

This matters because a truly acidifying fermentation can create conditions that inhibit certain pathogens.

A cooked food stored in a sealed jar without sufficient acidification has a different safety profile.

The word toroī identifies a traditional food and practice, but its use alone does not prove the final product has achieved a safe pH.

Ingredients — the 2007 research batch

Lorraine Louise Dixon's University of Waikato research prepared ten experimental batches.

Five used pūhā.

Five used watercress.

Each batch used one green rather than combining the two.

The combined batch weighed approximately 1.5 kg after processing.

The study used purchased mussels rather than relying on uncontrolled wild harvesting.

This is important because the researchers were investigating microbial safety and needed a defined starting material.

The exact edible mussel yield from two kilograms of shell-on mussels was not supplied separately.

The mussel shells therefore should not be counted as part of the edible finished batch.

The role of green-lipped mussels

The green-lipped mussel is a bivalve mollusc endemic to New Zealand waters.

It is an important seafood species and supports a substantial aquaculture industry.

Like other bivalves, mussels filter water while feeding.

This feeding behaviour allows them to accumulate microorganisms and other substances present in their environment.

Their safety therefore depends strongly on the quality of the waters from which they are harvested.

The fact that a mussel is alive when collected does not prove it is free of hazards.

Shellfish can accumulate pathogens or marine biotoxins.

Some hazards are not reliably eliminated by ordinary cooking.

For toroī, raw-material sourcing is therefore one of the most important safety considerations.

Pūhā and watercress

Pūhā is a wild edible green associated with Māori foodways.

Sow-thistle plants can have a bitter flavour.

Cooking changes their texture and can reduce some of the original bitterness.

Watercress is an aquatic plant with a peppery character.

Although both can be used in toroī, they are botanically different.

Pūhā belongs to the sunflower family, Asteraceae.

Watercress belongs to the cabbage family, Brassicaceae.

Their chemical compositions and environmental contamination risks therefore differ.

Watercress gathered from streams may be exposed to faecal contamination and other waterborne hazards.

The 2007 study used retail-sourced ingredients, not arbitrary wild plants.

The two variants should be recorded separately within the same recipe family.

Why the greens are cooked

The research method boils the selected green before combining it with mussels.

Cooking softens plant tissues and changes flavour.

It also reduces many vegetative microorganisms.

However, boiling does not necessarily eliminate all bacterial spores.

The botanical material may still introduce spores or other contamination during subsequent handling.

The research investigated Bacillus cereus, a spore-forming organism associated with various food materials.

The presence of a cooking stage is therefore relevant but does not establish sterilisation.

The vegetables also supply moisture and potentially fermentable compounds.

Their behaviour during storage can influence the final product.

The documented study preparation

The following stages reproduce the 2007 research process and must not be treated as a validated household preservation protocol.

1. Select one vegetable. Prepare either approximately 600 g pūhā or approximately 600 g watercress. The study used the two plants in separate experimental batches.

2. Clean the greens. Wash carefully, rub the stalks and remove dead leaves and large unsuitable stalks.

3. Cut the plant material. Prepare pieces approximately 5 cm long.

4. Boil the greens. Use 1 L tap water and boil for 30 minutes.

5. Drain. Retain the plant-cooking liquid for possible use. The amount subsequently required is not specified.

6. Clean the mussels. Scrub 2 kg shell-on green-lipped mussels under running water using a stiff brush.

7. Steam the mussels. The study uses a wire basket and steams until the shells open. It does not specify a measured internal cooking temperature or exact duration.

8. Remove the flesh. Use a sterile knife as described in the research protocol.

9. Chop the mussels. Cut the flesh into pieces approximately 1 cm long.

10. Reserve the liquid. Retain approximately 400 mL of the mussel-steaming liquid.

11. Combine and blend. Mix the cooked greens with mussel flesh and process to a purée.

12. Pack the experimental jars. The study used 70 mL and 120 mL screw-cap plastic jars with little headspace.

13. Store under research conditions. The prepared samples were held at 4°C and microbiologically assessed over time.

The exact fermentation endpoint is not established by these steps.

The storage periods in the thesis were research observation intervals rather than a universally safe household schedule.

Why the experimental vessel matters

The study used small plastic jars with screw caps.

This arrangement reduced exposure to external contamination but also affected oxygen availability.

A low-oxygen environment can favour certain microorganisms while inhibiting others.

For seafood products, that distinction has serious food-safety implications.

Clostridium botulinum is a particular concern in inadequately controlled, low-acid, reduced-oxygen seafood preparations.

The jar size, fill level, oxygen availability and storage temperature therefore matter.

However, the research did not establish these conditions as a complete validated safety system for arbitrary ingredients and storage periods.

A different vessel or larger batch cannot be assumed to behave identically.

What the microbiological research found

The 2007 thesis examined the product for several pathogens and investigated deliberate contamination challenges.

The experiments included Escherichia coli and Listeria monocytogenes.

The results demonstrated that refrigerated toroī could retain or support dangerous microorganisms under some conditions.

In the Listeria challenge, an inoculated strain increased approximately sevenfold over 19 days during refrigerated storage.

This is particularly important because refrigeration is often assumed to prevent bacterial growth completely.

Some pathogens, including Listeria monocytogenes, can grow at refrigeration temperatures.

The study also found that naturally present Bacillus cereus was detected early but was no longer recovered after two weeks in the investigated samples.

That observation does not establish that every future batch becomes free of B. cereus after the same period.

The study used particular raw materials, containers and analytical methods.

Research Commons

The E. coli investigation

A related study examined the survival of deliberately introduced E. coli in toroī.

Researchers prepared ten batches using cooked pūhā or watercress and green-lipped mussels.

They introduced E. coli at two different concentrations.

The initial levels were approximately 0.72 cells per gram and 604 cells per gram, expressed through the study's analytical method.

The contaminated samples showed declining E. coli numbers during storage, but the decline was slow.

The results demonstrated that refrigeration alone should not be relied upon to rapidly eliminate faecal contamination.

This is particularly relevant to shellfish and watercress harvested from contaminated environments.

Bacteriocins and Listeria control

The thesis also examined whether bacteriocin-producing microorganisms or purified bacteriocins could inhibit Listeria monocytogenes.

A strain of Lactobacillus sakei was investigated because it produces antimicrobial substances.

The researchers also studied nisin.

The experiments showed that particular interventions could reduce Listeria under specific conditions, but the effectiveness depended on the initial contamination and dose.

One study found that a L. sakei treatment could control a lower experimental inoculum but did not provide the same protection at a substantially higher contamination level.

These results are valuable scientific evidence.

They do not create a simple household recipe involving an arbitrary spoonful of starter culture.

The exact organism, amount and contamination conditions were part of a controlled research design.

The apparent contradiction in safety conclusions

The 2007 thesis abstract suggested that carefully prepared toroī could be safe after at least two weeks of refrigerated fermentation, provided ingredients were suitable and handling was carefully controlled.

However, the same research programme demonstrated growth of inoculated Listeria and slow disappearance of E. coli.

Earlier work also identified serious uncertainty over the protective mechanisms in traditional methods.

These findings should not be flattened into a blanket claim that two weeks at 4°C makes toroī safe.

The evidence instead shows that raw-material quality and contamination levels are critical, and that refrigeration alone is not a reliable pathogen-elimination step.

A public recipe page should describe the laboratory process without presenting it as an approved preservation method.

How toroī is eaten

Toroī may be served as a relish or soup.

The combination of mussels and greens creates a strongly marine, savoury preparation.

Mussel stock contributes flavour.

Pūhā or watercress adds a vegetal component.

The texture depends on whether the ingredients remain chopped or are blended.

The 2007 laboratory method produces a purée-like mixture.

Traditional household preparations may have a coarser texture.

The food can be understood within the broader Māori tradition of combining gathered marine resources and edible plants.

However, the source does not specify one universal serving temperature, portion size or compulsory accompaniment.

Those details should remain unknown rather than being invented.

Quality and safety

Toroī presents elevated food-safety concerns because it combines shellfish, moist vegetables and potentially prolonged storage.

Some versions rely on processes whose microbiological nature is not fully established.

The study's 4°C storage condition is a research parameter, not a comprehensive safety standard.

Shellfish sourcing must comply with relevant harvest-area controls.

Watercress should not be collected casually from potentially contaminated waterways.

Visible freshness, smell and the absence of mould cannot demonstrate that a batch is safe.

The correct editorial classification is traditional food and documented experimental process — not validated for unsupervised home preservation.