Authors: Entling, Martin; Ul Islam, A.S.M. Mufachcher;Bundschuh, Mirco;Gerken, Maximilian;Gerstle, Verena;Kolbenschlag, Sara L.;Manfrin, Alessandro;Brühl, Carsten A.;

In Europe, mosquitoes are widely controlled through applications of Bacillus thuringiensis var. israelensis (Bti) to wetlands. Although direct effects of Bti seem to be restricted to dipteran larvae, effects on non-target fauna are possible through trophic interactions. Field surveys reported conflicting results, partly finding no effects but sometimes observing widespread effects of Bti on dipteran predators including dragonflies (Odonata), spiders (Araneae) and birds (Aves). To test for a causal relationship between Bti and dragonfly populations, we set up a replicated field experiment with twelve artificial ponds mimicking permanent pools in river floodplains. Bti was applied to half of the ponds three times per year over two seasons. Dragonfly exuviae were collected, identified and prementum length measured for two species. Numbers of dipteran larvae in benthos samples was reduced in Bti-treated ponds, and emerging dragonflies were smaller, indicating food shortage. Numbers of emerging Libellulidae, but not of the larger-bodied Aeshnidae, were reduced in treated ponds, indicating negative effects of Bti through food shortage and/or intraguild predation. Our results demonstrate a conflict between Bit-based mosquito control and dragonfly conservation. Since Bti is often applied in protected areas, its environmental safety should be reconsidered.