Field trials were conducted to monitor the effects of very low doses (0.6-2 g a.i. ha(-1)) of fipronil on grasshoppers and non-target insects. The overall aim was to obtain efficient grasshopper control while inducing lower mortality among agriculturally beneficial organisms. The very low doses tested in Niger caused less short-term mortality in grasshoppers than higher doses (greater than 4 g a.i. ha(-1)) but very similar mortality rates were obtained in the long term. A 0.6 g a.i. ha(-1) spraying treatment was effective against grasshopper outbreaks, with 47% mortality obtained in 2 days and 91% in 10 days. Cage trials with grasshoppers showed that 1 and 2 g a.i. ha(-1) of fipronil were active for at least 23 days post-treatment. This acridicidal effect remained especially high during the first 8 days after spraying. Very low fipronil doses had an immediate impact on Coleoptera, Hymenoptera and Diptera. However, in most groups, the relative abundance of collected insects increased again in the last two survey samples, 20 and 32 days later. The side-effects on non-target invertebrate species were generally less severe and of shorter duration than with 4.2g and 13.4 g a.i. ha(-1) of fipronil, but comparable with the impacts of insecticides currently used in grasshopper control. In certain circumstances, low doses of fipronil could give good grasshopper control without any major threat to non-target invertebrates. (C) 1997 Elsevier Science Ltd.
