Differences in ecological impacts of systemic insecticides with different physicochemical properties on biocenosis of experimental paddy fields.

The environmental risks of pesticides are typically determined by laboratory single-species tests based on OECD test guidelines, even if biodiversity should also be taken into consideration. To evaluate how realistic these assessments are, ecological changes caused by the systemic insecticides imidacloprid and fipronil, which have different physicochemical properties, when applied at recommended commercial rates on rice fields were monitored using experimental paddy mesocosms. A total of 178 species were observed. There were no significant differences in abundance of crop arthropods among the experimental paddies. However, zooplankton, benthic and neuston communities in imidacloprid-treated field had significantly less abundance of species than control and fipronil fields. Significant differences in abundance of nekton community were also found between both insecticide-treated paddies and control. Influences on the growth of medaka fish were also found in both adults and their fries. Both Principal Response Curve analysis (PRC) and Detrended Correspondence Analysis (DCA) showed the time series variations in community structure among treatments, in particular for imidacloprid during the middle stage of the experimental period. These results show the ecological effect-concentrations (LOEC ~ 1 μg/l) of these insecticides in mesocosms, especially imidacloprid, are clearly different from their laboratory tests. We suggest that differences in the duration of the recovery process among groups of species are due to different physicochemical properties of the insecticides. Therefore, realistic prediction and assessment of pesticide effects at the community level should consider not only the sensitivity traits and interaction among species but also the differences in physicochemical characteristics of each pesticide.