Preference for plants damaged by conspecific larvae in ovipositing cabbage root flies: influence of stimuli from leaf surface and roots

In laboratory dual-choice assays females of the cabbage root fly, Delia radicum, prefer for oviposition plants with roots damaged by conspecific larvae to undamaged controls. Cauliflower and kale plants were inoculated with root fly eggs (25 per plant) and the hatching larvae were allowed to feed on the roots for various periods of time (1–17 days). After 4 (cauliflower) or 5 (kale) days of larval feeding the oviposition preference was most pronounced and flies laid between 64% and 68% of their eggs near plants with damaged roots. Later, with increasing damage but fewer surviving, and thus actively feeding, larvae, the magnitude of the preference declined. The preference for plants already damaged by conspecific larvae may contribute to the previously observed aggregated distribution of D. radicum eggs in Brassica crop fields.Further experiments revealed that the sensory cues inducing this oviposition preference originate from the complex consisting of the damaged roots, the surrounding substrate (soil) and associated microbes, rather than from the aerial plant parts. In choice assays using the root-substrate complex of damaged and control plants (aerial parts removed), the observed preference for damaged roots was similar to that found for the entire plant but was more pronounced. The damaged roots alone, compared to control roots, received up to 72% (cauliflower) and 75% (kale) of the eggs. By contrast, surrogate leaves sprayed with methanolic leaf surface extracts from the most preferred plants which had been damaged were not discriminated from surrogate leaved sprayed with extracts of the respective control plants. Analysis of glucosinolate levels in methanolic leaf surface extracts revealed that root damage resulted in enhanced concentrations of indole-glucosinolates on the leaf surface in kale but not in cauliflower. Although indole-glucosinolates are oviposition stimulants for the cabbage root fly, the induced changes were apparently too small to influence oviposition behaviour.