Laboratory mice injected with toxins from sea slugs found on an Auckland beach died as scientists watched - proving the slugs are toxic, but not revealing what has made them so poisonous.
Scientists trying to discover what killed two dogs that died after walking on Narrow Neck beach believe sea slugs found nearby may have eaten something toxic, possibly algae, and passed on the toxins to dogs.
But they are having to test for obscure chemicals, as tests for common toxins failed to find the culprit.
Cawthron Institute algae specialist Paul McNabb said the investigators made two important findings - that the slugs were toxic, and that the dogs' vomit was toxic.
Once they pinpoint the slug toxin, chemists will try to match it to the dog toxin. If the toxins are the same, they might be able to trace it to the ultimate source of the problem - whatever the slugs picked up in their environment.
Mr McNabb said chemists were trying to get "as far down the list as possible" of toxins by tomorrow, when health authorities are due to revise warnings against water sports and taking children to the beach.
If they had not identified a toxin by then, authorities would have to decide whether to approve "thousands of days" more of expensive tests.
Mr McNabb said the slugs were the only things taken from Narrow Neck beach found to be toxic, although seaweeds, sponges and a mixture of sand and algae were still being tested.
The stomach contents of pilchards, which have washed up in large numbers in recent weeks, were also being tested.
He said the toxic vomit strongly suggested the dogs died because of something they ate.
Councils and health authorities have stopped short of cancelling health alerts in the Hauraki Gulf until they know more.
A spokeswoman for the Auckland Regional Council, which is fronting efforts by eight Government and local agencies to find the cause of the problem, said further test results would be announced tomorrow afternoon.
The three grey sea slugs were collected from Narrow Neck beach last Friday but their bodies were in bits so biologists could not identify the species.
Mr McNabb said whole slugs would look like "shellfish without the shells" and would not necessarily be obvious to the naked eye if they were in dogs' vomit.
A dog would not need to eat a whole slug to become ill - picking one up in its mouth could be enough.
Unlike some tropical sea slugs, sea slugs found in New Zealand did not normally produce toxins, so it was more likely the slugs themselves had eaten something.
Niwa algae expert Dr Hoe Chang said sea slugs grazed on seaweeds and algal mats and could have been feeding on toxic algae.
But tests by him and the Cawthron Institute had not found any obvious species of toxic alga in the water.
Even shellfish, which feed on and collect toxic algae, did not test positive for any well-known problem alga.
Dr Chang said he initially thought an algal bloom was unlikely because most blooms in the past 20 years had been easily noticeable, usually turning the water brown.
There were no signs of a bloom this time, but algae could cause a toxic outbreak without reaching bloom levels.
A theory gaining currency among some amateur scientists - that the dogs were killed by fumes created when toxic algae are broken up by waves - was unlikely to be correct.
The species of alga that caused this type of effect, known as "toxic aerosols", had not shown up in any of the 21 water samples he had tested.
The worst-known effect of toxic aerosols was breathing difficulties, which disappeared once a dog or person left the water.
* The next steps
How scientists hope to find out what killed the dogs:
Identify the toxic chemicals in the sea slugs.
Prove the toxins are the same as those found in the dogs' stomachs.
Match toxins to something in the marine environment that the slugs have been eating.