Thèse Interaction Entre les Décisions Développementales et Comportementales de Recherche de Nourriture chez Caenorhabditis Elegans H/F - Doctorat.Gouv.Fr

Établissement : Université de Toulouse École doctorale : SEVAB - Sciences Ecologiques, Vétérinaires, Agronomiques et Bioingenieries Laboratoire de recherche : CRCA - Centre de Recherches sur la Cognition Animale Direction de la thèse : Audrey DUSSUTOUR ORCID 0000000213773550 Début de la thèse : 2026-10-01 Date limite de candidature : 2026-06-01T23:59:59 Ce projet de doctorat explore comment les décisions comportementales et développementales interagissent lors de la recherche de nourriture chez le nématode Caenorhabditis elegans. Bien que la théorie de l'optimisation du fourragement explique comment les organismes maximisent leur apport alimentaire, elle a principalement été appliquée à des animaux complexes. Chez les organismes plus simples, certaines hypothèses clés ne tiennent plus en raison de capacités sensorielles limitées et de cycles de vie rapides, où le développement et la prise de décision se déroulent sur des échelles de temps similaires.
Le projet vise à unifier les décisions comportementales, telles que quitter une source de nourriture, avec les choix développementaux, comme l'entrée en stade dauer - une forme résistante au stress déclenchée par des conditions défavorables. En utilisant C. elegans comme organisme modèle, nous combinerons des essais développementaux avec des expériences comportementales à haut débit afin de contrôler les variables environnementales et de suivre le comportement individuel.
Nous testerons comment les expériences passées influencent les trajectoires développementales et si des principes communs de prise de décision régissent à la fois le comportement et le développement. À terme, nous visons à construire un cadre unifié expliquant comment les organismes intègrent l'information au cours du temps pour optimiser leurs stratégies de survie.
Obtaining food is not only key for survival; it is one of the most time-consuming activities for many organisms. For this reason, even minute efficiency gains can provide a significant advantage and foraging must be highly optimized. This high degree of optimization, together with the fact that its outcome is relatively easy to measure (in terms of the amount of food collected per unit time) makes foraging an ideal test-bed to understand how organisms make decisions and execute them. For this reason, foraging is deeply studied, both from a theoretical and from an experimental point of view1. In particular, Optimal Foraging Theory studies the optimal strategies that animals should follow to maximize their food intake1-3.
However, most foraging studies focus on relatively complex animals, from insects upwards1,3. Simpler organisms, from bacteria to lower invertebrates, comprise a great part of the tree of life, and have characteristics that break some of the assumptions of classical Optimal Foraging Theory. These characteristics belong to two main classes:
First, simple organisms have strong sensory and cognitive constraints (in particular, they have no vision). While Optimal Foraging Theory considers these types of constraints1, and substantial work has been done in several areas such as navigation and chemotaxis4,5. However, other areas, such as optimal food exploitation, are underexplored.
The second characteristic of simple organisms is their fast life cycle, which can be as short as a few minutes for bacteria, and of a few hours or a few days for many other simple organisms. This short life cycle breaks one of the implicit assumptions of Optimal Foraging Theory: The separation of timescales between decisions and development. This separation of timescales happens when decisions have a timescale much shorter than the individual's lifespan. For example, the decision of leaving a food patch has consequences (e.g. finding a different food patch) that happen in a timescale of, at most, hours. An animal with a lifespan of months has a good separation of timescales, because it changes very little developmentally during those hours. In contrast, for many simple organisms there is poor separation of timescales between decisions and development: An organism with a life cycle of one day will grow significantly in a few hours, often undergoing important developmental changes, so the individual that suffers the consequences of a decision is often very different from the individual that made the decision.
Our laboratory is addressing all these characteristics to adapt Optimal Foraging theory to simple organisms. First, we are developing Optimal Foraging models specifically tailored to the constraints of non-visual organisms6. Second, we study the interplay between behavioral and developmental decisions. This thesis will focus on this second domain, benefitting from the lab's expertise in the overall question. This project will use the nematode Caenorhabditis elegans as an experimental model system. C. elegans is a perfect model for this project: It is a perfect example of a simple organism, having a fast developmental cycle (48 hours egg-to-egg), having no vision and having a small nervous system (around 300 neurons). It is also experimentally tractable, and very well studied.7
C. elegans has a life cycle shaped by fluctuating environmental conditions. In nature, it experiences boom-and-bust dynamics on ephemeral food sources: The population expands rapidly when resources are abundant, and collapses when food runs out. To survive starvation, newborn L1 larvae can follow two alternative developmental trajectories: they may proceed directly to reproductive adulthood under favorable conditions, or enter the dauer stage, a stress-resistant and long-lived larval form. Dauers can survive harsh environments and resume normal development once they encounter improved conditions, such as renewed food availability. Thus, the dauer decision represents a critical developmental checkpoint that integrates environmental cues to determine life history strategy.
Our laboratory is developing a framework that uses Foraging Theory to unify these developmental decisions with behavioral foraging decisions7 (Figure 1). This framework describes the decision to become dauer as equivalent to leaving a food source, and can therefore apply the same principles that govern the behavioral decision to leave a patch to the developmental decision to become dauer larva. In this framework, information about the quality of the environment is collected at all life stages, and even transmitted trans-generationally, and used to inform both behavioral and developmental decisions, according to optimal decision-making theory. 7
This project will test this framework experimentally, focusing on the interplay between behavioral decisions (when to stay in a food patch and when to leave it) and developmental decisions (whether to develop into dauer larva or into adult). These two decisions are well studied separately, but never together. In particular, food choice and patch-leaving decisions are well studied in C. elegans adults8-10, but no published data exists about these decisions in any other larval stage. As for developmental decisions, they have only been studied in conditions with the amount of food and pheromone were fixed, so the worm's behavior had very little impact on the outcome11. This project will study for the first time the interplay between these two types of decisions, by studying the developmental Dauer decision in conditions where the worm's behavior plays an important role.

We will combine two types of experimental methods: Developmental and behavioral assays.
Developmental assays
In these experiments, C. elegans individuals develop in conditions where the density of food and pheromones is carefully controlled, and the resulting larvae develop during 4 to 5 days, either becoming adults or becoming dauer larvae. We can thus study what concentrations of food and pheromone induce dauer formation, and how this decision depends on the previous experience of the individual or of its ancestry. Our laboratory has extensive experience with these assays.
Behavioral assays
We have built a high-throughput experimental pipeline that includes a customized pipetting robot to prepare large numbers of experimental arenas with high reproducibility and high flexibility (up to 600 arenas per day, with any arbitrary distribution of food patches and/or sources of chemoattractants). The robot automatically randomizes and labels each experimental condition, which turned out to be one of the main bottlenecks of the process. We have three types of imaging devices to visualize the worms (Figure 2): (1) Scanners, which provide a single snapshot of the positions of worms. (2) A low-resolution multi-camera setup, which records 48 arenas simultaneously at 1 frame per second, and (3) A high-resolution multi-camera setup, which records 6 arenas simultaneously at 12.5 frames per second.

Le candidat ou la candidate doit avoir une expérience dans la réalisation d'expériences comportementales avec Caenorhabditis elegans ainsi que dans l'analyse de données comportementales.