The Department of Conservation (DoC) is planning to upset the stomachs of kea that scavenge for food from tourists - a move designed to save the birds' lives.
Tourists frequently ignore signs asking them not to feed the cheeky South Island native parrots. And by doing so, they encourage kea to try new foods.
That is exactly what DoC does not want kea to do, as it embarks on its ''Battle for our Birds'' - a programme to spread cereal pellets laced with 1080 across 500,000ha of native beech forest in spring this year.
The poison is aimed at rats, mice, possums and stoats that attack the eggs and chicks of native birds.
But kea, more so than other native birds, are prone to trying the lethal pellets.
DoC science adviser Josh Kemp is involved in a programme to find a way of discouraging them.
"If you sit down and draw up a list of behavioural traits and characteristics that would put a bird at risk, like feeding on the ground, being inquisitive, being an omnivore ... you end up with keas and wekas at the top."
Mr Kemp told the Otago Daily Times this week the work was not going as "swimmingly as we had hoped" but it had not gone "completely badly either".
Field trials near Arthurs Pass last year using bait laced with both 1080 poison and the repellent d-pulegone resulted in five kea deaths from among 39 birds being monitored.
Mr Kemp suspected the reason the birds ate the bait and died was because they had been encouraged by tourists to try new things.
"Keas that have had a history of scavenging at car parks and alpine villages may be more likely to die from 1080 poisoning.
"They have got into the habit of investigating novel things when they are hanging around car parks, looking through rubbish bins and trying things people are throwing to them."
Asked what evidence there was to support the link between tourists feeding kea and deaths from 1080, Mr Kemp said it was "just a theory", based on the results of the field trial in which kea had died.
"It's a really hard theory to prove. We can't do an experiment in a lab, like you might like to do, to actually really get a handle on what's going on.
"But there is quite a strong likelihood ... people are making things worse. It's a theory we can't reject at all."
Mr Kemp said tourists fed kea to amuse their children or to get photos and often believed they were the only ones doing it.
"People just can't resist it."
Policing car parks would be difficult and expensive, Mr Kemp said, and instead, DoC was considering the use of bird-repellent anthraquinones to "train" scavenging kea.
The chemical causes a mild physiological reaction - "a gut irritation, basically, that makes them feel crook" - within a few minutes of them eating a pellet. The kea would associate the upset with eating the pellet.
"So we are looking at doing that at all the skifield car parks and other car parks and villages."
Mr Kemp said "preliminary observations" on captive birds suggested it would work.
"The birds that have taken a good mouthful of this stuff, or a good belly full of this stuff, have exhibited signs of temporary discomfort and been really put off the pellets."
Anthraquinone is used on crop seeds to deter birds.
Asked if feeding chemicals to wild kea to make them sick might upset the public, Mr Kemp said the idea was better than seeing the birds dying from 1080 poisoning.
Mr Kemp said kea numbers were "going downhill" in areas where there was no predator control.
In his experience, kea in the "back blocks" of the beech forests that had little human contact tended to ignore unnatural "junk food" and there was "zero" evidence of those particular kea dying from eating 1080 pellets.
For that reason, 1080 pellets dropped in beech forests this year would not contain a repellent.
The main problem area was the large rimu, rata and kamahi forests of the West Coast, where there was much tourist activity.