The role of glucosinolates in the oviposition behaviour of the cabbage root fly, Delia radicum (L.) (Diptera, Anthomyiidae) was investigated using egg counts and electrophysiological recordings from tarsal contact chemoreceptors. The glucosinolates present both inside and on the surface of cauliflower leaves were determined. The total amounts obtained with the two methods differed by a factor of 100. The extract of the leaf surface contained about 60 μg per g leaf extracted (gle), the total leaf extract 7.5 mg per gle. The glucosinolate patterns of the two extracts were qualitatively similar, but the ratios of the content of individual glucosinolates showed considerable differences. The D sensilla on segment 3 and 4 of the tarsus of D. radicum females were shown to contain a sensitive receptor cell for glucosinolates. In contrast, the receptor cells of the D sensilla of the other segments did not respond in a dose dependent way to these compounds. The glucosinolate receptors were found to be especially sensitive to glucobrassicin, gluconasturtiin and glucobrassicanapin with thresholds of about 10 −8M to 10 −9M. Large differences (up to two orders of magnitude) were observed among the different glucosinolates. A significant correlation was found between the behavioural discrimination index and the electrophysiological results. But no obvious correlation existed between the chemical nature of the glucosinolate side chain (e.g. indole, aromatic and aliphatic groups), and their stimulatory activity. However, a significant correlation was found between the overall length of the side chain and the biological activity. Although the flies discriminated clearly between model leaves with and without glucosinolates, a clear dose response curve was only obtained for the indole glucosinolate glucobrassicin. Since the most stimulatory fraction of the surface extract contained no glucosinolates, it was concluded that other compounds, in addition to glucosinolates, do play an important role for the stimulation of oviposition.

The time-course of behavioural phase change was investigated in nymphs of Schistocerca gregaria, using logistic regression analysis of behaviour recorded in a standard assay. Gregarization occurred very rapidly. Solitaryreared nymphs became markedly gregarious in behaviour within 1-4 h of being placed in a crowd. These insects re-solitarized equally quickly if removed from the crowd. Crowd-reared locusts also solitarized within 1-4 h, but this effect was not complete. Results indicate that, while behavioural gregarization is maximal within a few hours of crowding, solitarization is a two-stage process, changing rapidly at first, then more slowly as a function of the period of previous crowding.

Abstract. Contact chemoreception plays a decisive role in host selection and oviposition behaviour of the cabbage root fly, Delia radicum L. (Diptera, Anthomyiidae). Glucosinolates (mustard oil glucosides) are known to be perceived by the flies, and when sprayed on paper leaf‐models induce oviposition. Recently it has become clear that other non‐volatile types of compounds must also be involved in host selection. A pair of ventro‐medial C sensilla on die fifth tarsomere respond strongly to a novel compound called tentatively ‘cabbage identification factor’ (CIF), but not to sucrose, glucose, fructose and proline. CDF is a new non‐glucosinolate oviposition stimulant. A single neurone in each sensillum is activated by this compound and the same is true for glucosinolates. In some flies a mixture of bom types of stimuli evoked an apparent mononeural spike train, whereas in odiers spikes of two separate cells were activated. The significance of this variability is not yet clear. The new stimulant, CIF, does not evoke responses in glucosinolate receptors in the D sensilla. The involvement of the C3 sensilla in the detection of host‐specific compounds constitutes the first known function for C sensilla in D. radicum. CIF appears to be present in leaf surface extracts from the host‐plant Brassica oleracea in quantities as low as 1 ng per gram leaf. In spite of this low level, it stimulates oviposition significantly better than glucobrassicin at higher concentrations, which up till now was known as the most powerful stimulant for D.radicum.

The nature of stimuli, emanating from other locusts, which are effective in inducing gregarization in the desert locust was investigated. Isolated-reared fifth-instar nymphs were subjected to tactile, visual and olfactory stimuli, presented singly and in combination, and the effect on the behavioural phase state was quantified using logistic regression analysis. Tactile stimulation provided by rolling paper spheres proved to be highly gregarizing, whether presented alone or in combination with the other stimuli. Olfactory and visual stimuli together caused partial behavioural gregarization. Visual stimulation alone was weakly gregarizing after prolonged exposure, while olfactory stimuli alone were ineffective. Nymphs and pre-reproductive and reproductive adults of both sexes were also treated with synthetic adult male `aggregation' pheromone blend (Torto et al., 1994, Journal of Chemical Ecology 20, 1749). No effect of this blend was found on the behavioural phase state, even when visual stimuli were present. Non-locust related stimuli, including wheat odour and flashing lights, were also tested on nymphs. Neither induced any change in the behavioural phase state, indicating that increased sensory flow is not a sufficient explanation for locust-induced behavioural phase change.