Abstract
Phase polyphenism is an extreme form of density-dependent phenotypic plasticity expressed by about 20 species of grasshoppers of the Acrididae family, named locusts. Locusts present two extreme phenotypes: the "solitarious phase" and the "gregarious phase", showing very different characteristics. The two phases are distinguished in particular by their way of life: solitarious locusts will adopt a sedentary and cryptic behaviour, while the gregarious ones will gather and move in a coordinated way over long distances, as devastating bands of nymph or swarms of imagos. In this thesis, we focus on this particular aspect of phase polyphenism: the collective movements of gregarious individuals, which we will study under different spatio-temporal scales.The first part of this thesis aims to explore the impact of spatial variability of resources on collective foraging at a short spatiotemporal scale, corresponding to the distance covered by a gregarious nymphs band in a few hours of walking (< 100 m). For this purpose, we developed an agent-based model, allowing to represent local interactions between individuals and with vegetation, at a temporal scale of 10s. On a large variety of landscapes, we studied under which conditions the group will have an advantage over solitarious nymphs in terms of foraging. Our results highlight the importance of alignment within the band to optimize foraging, and show that certain landscapes where the resource is aggregated and sparse (occupying less than 40% of space) favor gregarious individuals.A second part explores the hypothesis of an attractive effect of faeces on gregarious locusts, which could notably allow gregarious nymphs that got lost to easily find the group's trace. Such an effect could ensure a better cohesion of the group on a larger spatio-temporal scale (of the order of the day, for a distance higher than 100m). To this end, we performed behavioural olfactory assays in the laboratory on L3-stage nymphs of the desert locust, Schistocerca gregaria, in the presence of faeces from several age classes (1h or 24h). We also performed chemical analyses (GC-MS) to explore which volatile organic compounds were emitted from nymphs faeces. Our results show an attractive effect of the 1h and 24h faeces on the nymphs, suggesting that the effect could last for at least one day and thus allow latecomers individuals to find the group's trace.Finally, we discuss the possible implications of these results in the context of the emergence of locusts' phase polyphenism, and prospects for future studies on the subject at higher spatiotemporal scales. Given our modeling results, and knowing that a variable environment favors the emergence of phenotypic plasticity, resource variability could be one of the factors favoring the emergence of phase polyphenism. Evolutionary optimum models could demonstrate the benefits of polyphenism, allowing to live isolated or in groups according to the variability of the resources. Demogenetic agent-based models would allow to study the influence of spatial and temporal variability of the resource on the emergence and evolution of phase polyphenism, through the locust gregarization threshold. Taking into account the attractive effect of feces, ensuring better group cohesion, could have implications on the results of such a model. This future work would test whether the evolution of phase polyphenism results from interactions at varying spatiotemporal scales.