Prof John Mumford, Professor of Natural Resource Management at Imperial College London, said:

"It would also be prudent in a risk analysis to seek evidence of the levels of tetracycline in the environment that would be likely to be encountered by released mosquitoes and their offspring. We could well imagine that rearing mosquitoes with treated chicken could lead to exposure to tetracycline, but what happens across a range of real breeding grounds? If there are points of high tetracycline concentration then the risk analysis would need to consider a risk management measure that would deal effectively with it, for example localised insecticide treatment in the vicinity of high tetracycline points.

"Allergenicity is raised and indeed a risk analysis should identify that as a hazard and attempt to estimate the likelihood, related to the possible number of females arising from contamination or unintended survival of offspring. I would want to set that against health risks from unintended dengue cases arising from the alternative baseline actions, with attendant frequency and impacts.

"I have not been part of the risk analysis process in any of the countries where RIDL mosquitoes have been released so far, so I cannot say if the points above have been addressed by the authorities.

"In the case of RIDL diamondback moth in the UK, Defra/HSE have been advised to seek evidence related to the fate of any levels of female and non-sterile releases arising within the proposed operations. Allergenicity in the field is not an issue for diamondback moth, but may be in rearing facilities, where physical contact can be relatively high.

"The answers in risk analyses may well be that the hazards have been identified and the risks are either acceptable or can be reliably managed to be acceptable."