Desert Locusts can change reversibly between solitarious and gregarious phases, which differ considerably in behaviour, morphology and physiology. The two phases show many behavioural differences including both overall levels of activity and the degree to which they are attracted or repulsed by conspecifics. Solitarious locusts perform infrequent bouts of locomotion characterised by a slow walking pace, groom infrequently and actively avoid other locusts. Gregarious locusts are highly active with a rapid walking pace, groom frequently and are attracted to conspecifics forming cohesive migratory bands as nymphs and/or flying swarms as adults. The sole factor driving the onset of gregarization is the presence of conspecifics. In several previous studies concerned with the mechanism underlying this transformation we have used an aggregate measure of behavioural phase state, Pgreg, derived from logistic regression analysis, which combines and weights several behavioural variables to characterise solitarious and gregarious behaviour. Using this approach we have analysed the time course of behavioural change, the stimuli that induce gregarization and the key role of serotonin in mediating the transformation. Following a recent critique that suggested that using Pgreg may confound changes in general activity with genuine gregarization we have performed a meta-analysis examining the time course of change in the individual behaviours that we use to generate Pgreg. We show that the forced crowding of solitarious locusts, tactile stimulation of the hind femora, and the short-term application of serotonin each induce concerted changes in not only locomotion-related variables but also grooming frequency and attraction to other locusts towards those characteristic of long-term gregarious locusts. This extensive meta-analysis supports and extends our previous conclusions that solitarious locusts undergo a rapid behavioural gregarization upon receiving appropriate stimulation for a few hours that is mediated by serotonin, at the end of which their behaviour is largely indistinguishable from locusts that have been in the gregarious phase their entire lives.

Nutritional regulation is a powerful mechanism used by generalist feeders to obtain the balance of nutrients they require from nutritionally diverse, perhaps unbalanced, foods. We examined nutritional regulation in a species with a narrow individual diet breadth: the forest tent caterpillar, Malacosoma disstria. Fourth instar caterpillars were provided with artificial foods consisting of different ratios of protein to digestible carbohydrate in no-choice, choice and compensatory feeding experiments. In the no-choice test, caterpillars were confined to a single food source of varying protein/carbohydrate ratio for the duration of the fourth larval stadium. Caterpillars performed best on equal-ratio and slightly protein-biased diets. Significant reductions in performance were only observed on extremely protein- or carbohydrate-biased diets. Daily consumption of the three acceptable intermediate diets was consistent with volumetric regulation, but the timing of the moult to the next instar appeared linked instead to protein intake. In the choice test, caterpillars were provided with two complementary foods, one biased toward protein and the other toward carbohydrate, for the duration of the stadium. The caterpillars fed randomly from the two food sources presented to them, except for the extremely protein-biased diet (P:C ratio of 35:7), which they avoided. The compensatory feeding experiment tested whether forest tent caterpillars deprived of either protein or digestible carbohydrate would select a food containing the deficient nutrient. Insects were conditioned on either protein-only, carbohydrate-only, protein-and-carbohydrate or no-nutrient foods, then offered a choice between protein-only and carbohydrate-only foods. Unlike previously studied generalist feeders, our caterpillars did not compensate for protein deficiency and showed only very weak evidence of compensation for carbohydrate deficiency. Forest tent caterpillars are colonial trail-laying forest folivores that are generally confined to a single host plant and hence do not experience much diversity in food nutrient ratios. We show that forest tent caterpillars do not independently regulate protein and carbohydrate intake. These findings are consistent with predictions that nutritional regulation abilities should be less important in animals with narrower diet breadths.

Synchronized behaviour is common in animal groups. In ant colonies, synchronization occurs because active ants stimulate their neighbours to activity. We use oscillator theory to explain how stimulation from active neighbours synchronizes activity in groups of solitarious locusts via entrainment of internal physiological rhythms. We also show that the spatial distribution of food resources controls coupling between individual locusts and the emergence of synchronized activity. In locusts ( Schistocerca gregaria ), individual schedules of activity and quiescence arise from an irregular physiological oscillation in feeding excitation (i.e. hunger). We show that contact with an active neighbour increases the probability that a locust becomes active. This entrained activity decreases the time until the locust feeds, shifting the phase of its hunger oscillation. The locusts' internal physiological rhythms are thus brought into alignment and their activity becomes synchronized. When food resources are clumped, contact with active locusts increases, and this increase in the strength of coupling between individuals leads to greater synchronization of behaviour. Activity synchronization might have functional significance in inhibiting swarming when resources are dispersed and accelerating it in more favourable clumped environments.