Genetic variability in the area searched by a parasitic wasp: analysis from automatic video tracking of the walking path

The ability of foraging hymenopterous parasitoid females to discover their hosts, and thus to be efficient agents in biological control programs, is likely to be related to the surface they are able to prospect per unit of time. However, this behavioural trait has never been accurately estimated, and its implication in female efficiency has never been really demonstrated. This paper provides an algorithmic method that can be used to estimate this trait from an automatic recording of the females' walking path. Using stochastic procedures simulating walking tracks, this trait is shown to be strongly related to the number of hosts that parasitoid females are able to attack per unit of time. This trait was estimated for individual females of Trichogramma brassicae Bezdenko (Hymenoptera; Trichogrammatidae). On average, females of this species are able to prospect about 28mm(2)s(-1). Finally, the genetic variation in this trait was studied using the iso-female line method. A significant genetic variability was observed. It provides the basic information that is necessary to start a genetic selection of mass-reared Trichogramma in order to improve their efficiency in controlling target pests in biological control programs. The functional and evolutionary implications of these results are discussed.