AgResearch senior scientist Craig Phillips said a small number of clover root weevils had been found on one farm at Clinton, and scientists had been experimenting with how to release the biological control, a parasitic wasp, without releasing more weevils.
The wasp was traditionally spread by releasing weevils infected with wasp larvae, but Prof Phillips said there was no guarantee they all carried the larvae, meaning healthy weevils could inadvertently be released.
This was a risk in South Otago where the weevils were discovered in low numbers on one farm, but were expected to have spread to others.
Prof Phillips said AgResearch had trialled a new system for releasing the biological control wasp at weevil-infected sites in Golden Bay and Canterbury, where weevils infected with the parasitic wasp were kept in boxes in the field.
Very fine mesh on the sides of the boxes allowed the parasitic wasp, once hatched and mature, to escape and attack weevils.
The mesh size meant weevils stayed contained.
Prof Phillips said the system worked in the laboratory and was being proven in the field, and could be extended to Clinton in late spring or early summer.
He said scientists were waiting for weather data to confirm when to release the wasps.
The wasp was proving to be more hardy and effective in combating weevils than scientists had initially hoped.
"It's managed to survive very difficult drought conditions in the North Island, on the East Coast and Waikato when clover disappeared and the CRW [clover root weevils] also disappeared," he said.
The weevils had also proven very adaptive after being established in the North Island, then found at Christchurch Airport in 2006 before spreading 20km northwest of the airport and into Nelson.
There had been no signs of the weevils south of Christchurch until their discovery at Clinton last year.
Given their ability to travel on the wind, Prof Phillips said he expected to find new populations elsewhere in Clinton.
Farmers would see the first signs of infestation on clover with notches appearing on the clover leaf.
This would be followed by an eventual decline in the performance of clover.