Japan's agriculture ministry has opened a public comment on a draft assessment of dinotefuran, a neonicotinoid insecticide, and its effect on honeybees. The draft concludes that the chemical is unlikely to impair the maintenance of bee colonies, but only if it is used under the applied-for methods and risk-management measures. Comments are accepted from October 15 to November 13, 2026, and postal submissions must arrive by the deadline. The ministry describes the exercise as a voluntary consultation, not a procedure required under the Administrative Procedure Act.
What the draft concludes
The Ministry of Agriculture, Forestry and Fisheries (MAFF) says the Pesticide Control Act requires the minister to seek the Agricultural Materials Council's opinion on registrations and changes. Honeybee questions go to a specialist subcommittee, which approved this draft at its 22nd meeting on September 10, 2026. The draft covers 93 formulations and names the applicant. MAFF says it will decide the final text after weighing the comments it receives.
How the draft gets there
The draft sets four reference toxicity values for bees.
| Life stage and exposure | Test | Value |
|---|---|---|
| Adult, single contact | 48 h LD50 | 0.062 µg ai/bee |
| Adult, single oral | 48 h LD50 | 0.013 µg ai/bee |
| Adult, repeated oral | 10 d LDD50 | 0.010 µg ai/bee/day |
| Larva, oral | 72 h LD50 | 3.3 µg ai/bee |
Because the adult contact and oral LD50 values are below 11 µg per bee, the draft says a caution statement is required. The first-stage test treats mortality of up to 10% as acceptable, and uses a risk quotient of 0.4 or less as its proxy for that threshold.
Some uses are judged to involve no bee exposure at all, including small-volume products and crops harvested before flowering, such as cabbage and lettuce, or crops that do not flower, such as sweet potato and tea. Others are allowed only with conditions, such as use in closed-system facilities or after flowering ends.
Where the limits bite
In wasabi seedling-stage spraying, the screening risk quotients exceeded 0.4. The draft refined them with residue data from pumpkin, used as a representative crop, and the quotients fell to 0.4 or below. For soil treatments on cucumber, melon, chili peppers and similar crops, refinement also brought every use to 0.4 or below, though peppers and chili types need use-quantity limits.
Rice and forage grass went to a second-stage colony test. Two US tunnel trials found no effect on adult dead-bee counts or colony strength, according to the draft. In the first, however, dead-bee counts inside the tunnel differed significantly from the control, with a calculated death rate of 1.8%. For forage grass, the draft adds a condition: where more than 150 litres per 10 ares is used, the crop must be harvested before flowering.
The draft set aside studies of behavioural effects such as homing flight and olfactory learning. It says they do not show a serious effect on colony maintenance and that no link to colony-level harm is known, but adds that new findings on behavioural abnormalities must keep being tracked. The conclusion is confined to colony maintenance in a honeybee impact assessment.
How to comment
Submissions must be in Japanese and include a name, address and contact details. Comments may be published, with personal information withheld. MAFF takes no comments by phone and will not reply to individuals.