Get access

Genome scan in the mosquito Aedes rusticus: population structure and detection of positive selection after insecticide treatment


Margot Paris, Fax: +33 4 76 51 42 79; E-mail:


Identification of genes involved in local adaptation is particularly challenging for species functioning as a network of interconnected populations undergoing frequent extinctions–recolonizations, because populations are submitted to contrasted evolutionary pressures. Using amplified fragment length polymorphism markers, population genetic structure of the mosquito Aedes rusticus was analysed in five geographical areas of the French Rhône-Alpes region. We included a number of sites that were treated with the bio-insecticide Bacillus thuringiensis israelensis (Bti) for more than 15 years. Analysis of molecular variance revealed that most of the genetic variability was found within populations (96%), with no significant variation among geographical areas, although variation among populations within areas (4%) was significant. The global genetic differentiation index FST was low (0.0366 ± 0.167). However, pairwise FST values were significant and no isolation-by-distance at the regional level was observed, suggesting a metapopulation structure in this species. Bti-treatment had no effect on genetic structure and on within-population genetic diversity. Potential signatures of positive selection associated with Bti-treatment were detected for five loci, even though toxicological bioassays performed on field-collected larvae showed no significant difference in mortality between Bti-treated and nontreated sites. The difficulty of detecting moderate resistance in field-collected larvae together with possible differential persistence of toxins in the environment may explain our inability to detect a toxicological response to Bti in treated sites. The evidence for positive selection occurring at several genomic regions suggests a first step towards Bti resistance in natural mosquito populations treated with this bio-insecticide. Furthermore, this signal was detectable using genomic tools before any toxicological evidence for resistance could be identified.