Long-term monitoring reveals among-year consistency in the ecological impacts of insecticides on animal communities in paddies

Abstract

Abstract Semi-natural mesocosm experiments have been proposed as a powerful approach for evaluating the realised rather than potential ecotoxicological effects of agrochemicals on biological communities. However, there is uncertainty in the results from such experiments due to spatio-temporal variability in various ecological factors, such as community composition. Therefore, long-term studies evaluating such variability are critical for assessing the utility of mesocosm experiments, although few studies have been conducted over multiple years. We conducted a multiple-year (i.e., three-year) mesocosm experiment to examine the effects of three systemic insecticides (clothianidin, fipronil, and chlorantraniliprole) on aquatic animal communities in experimental paddies, and we evaluated the among-year variability in these ecotoxicological effects. We found substantial variation in community composition among years. However, when we controlled for this variability by performing partial redundancy analysis (partial RDA), we found that paddy community composition was significantly different between the three insecticide treatments irrespective of year. Moreover, we identified some taxa that consistently decreased in abundance following exposure to each insecticide and may thus be regarded as indicator taxa of the environmental stress caused by the insecticides (e.g., clothianidin: Gerris lacustris (water striders), fipronil: Orthetrum albistylum speciosum (dragonflies), and chlorantraniliprole: Hydroglyphus japonicus (beetles)). Our study suggests the utility of multiple-year, semi-natural mesocosm experiments not only for evaluating the temporal variability in ecotoxicological effects but also for identifying potential robust indicator organisms that consistently decrease in abundance in response to specific agrochemicals irrespective of the variability in ecological factors under realistic conditions.