Phylogeography and population genetics of swarming grasshoppers

Population studies focusing on neutral sequence variation are recurrently cited as promising candidates to assess the population structure and dynamics of outbreaking insect populations. These questions remain, however, largely unstudied in locusts, in part because of the difficulty of getting suitable molecular markers at both phylogeography and population genetics levels. Undetected intra-individual variation in mitochondrial DNA (as well as in nuclear ITS) sequence data is particularly common in grasshoppers and can confound phylogeographic analyses. The popular and versatile microsatellite loci are prone to both (i) multiple banding amplification patterns due to high similarity between their flanking regions and (ii) high prevalence of null alleles in Orthopteran species. In order to better develop operational microsatellite markers in locusts, we studied the distribution patterns and the molecular evolution of the similar microsatellite-containing sequences repeated within the genome both within and among species of the Orthoptera. In order to comprehensively describe the genetic variation within and among populations of locusts, we assessed null allele effects on statistics and methods traditionally used to analyse microsatellite variation and, in some cases, developed adapted statistical tools for genotype analysis. These works enabled us to build a comprehensive picture of the population genetic structure of two major locust species, the cosmopolitan Locusta migratoria, and the Australian-endemic Chortoicetes terminifera. Firstly, we assessed the role of the propensity to outbreak in shaping the worldwide and regional patterns of population genetic variation in L. migratoria. Secondly, we showed an absence of genetic differentiation between C. terminifera populations from Western Australia and the four Eastern states of Australia, which have been monitored independently, largely assuming low connectivity between populations from both sides of the continent. We used computer simulations and literature survey with the aim to better consider the implications of our molecular genetics findings with respect to spatial population dynamics and management of locusts. (Texte integral)