Red flag for green spray: adverse trophic effects of Bti on breeding birds

Authors


Correspondence author. E-mail: poulin@tourduvalat.org

Summary

1. The expanding use of selective pest-control agents provides a unique opportunity to study food web interactions in the field while addressing major environmental issues. Bacillus thuringiensis israelensis (Bti) is the most commonly used microbial agent to control mosquitoes worldwide.

2. Using breeding house martins Delichon urbicum as a model species, we assessed the effect of Bti spraying on foraging rates and chick diet prior to and during 3 years of Bti spraying in the Camargue, France. Some 9051 feeding flights and 14 857 prey items were recorded in the early, mid and late nesting season at up to three control and three treated sites. Breeding parameters were assessed during 1 year at two control and two treated sites.

3. Intake of Nematocera (Diptera sub-order including midges and mosquitoes) and their predators (spiders and dragonflies) decreased significantly at treated sites, concurrently with increase of flying ant intake. Small prey (<2·5 mm) were significantly more taken at treated sites, and large prey (>7·5 mm) at control sites, with lower foraging rates at treated sites.

4. Clutch size and fledgling survival were significantly lower at treated sites relative to control with respectively 2·3 vs. 3·2 chicks produced per nest. Breeding success was positively correlated with intake of Nematocera and their predators at the nest level.

5. No previous study has provided compelling evidence of Bti affecting vertebrate populations following the suppression of prey species. Indirect effects caused by repeated application of Bti through food web interactions warrant more attention.

6.Synthesis and applications. Bti is considered the most selective and least toxic agent currently available to control mosquitoes. Mosquito-control programmes should integrate non-biased awareness campaigns and mitigation measures balancing the social demands for mosquito reduction with the factors involved in mosquito proliferation and dispersion. Such measures could consist in improved wetland management; reduction in areas and periods of Bti spraying; consideration of alternatives to Bti spraying, such as mosquito traps; specific measures to reinforce animal populations affected by Bti; and suspension of mosquito control in environmentally sensitive areas where nature preservation is a priority.

Ancillary