A neuronal substrate for translating nutrient state and resource density estimations into foraging decisions

Abstract

Foraging animals must balance the costs of exploring their surroundings with the potential benefits of finding nutritional resources. Each time an animal encounters a food source it must decide whether to initiate feeding or continue searching for potentially better options. Experimental evidence and patch foraging models predict that this decision depends on both nutritional state and the density of available resources in the environment. How the brain integrates such internal and external states to adapt the so-called exploration-exploitation trade-off remains poorly understood. We use video-based tracking to show that Drosophila regulates the decision to engage with food patches based on nutritional state and travel time between food patches, the latter being a measure of food patch density in the environment. To uncover the neuronal basis of this decision process, we performed a neurogenetic silencing screen of more than 400 genetic driver lines with sparse expression patterns in the fly brain. We identified a population of neurons in the central complex that acts as a key regulator of the decision to engage with a food patch. We show that manipulating the activity of these neurons alters the probability to engage, that their activity is modulated by the protein state of the animal, and that silencing these neurons perturbs the ability of the animal to adjust foraging decisions to the flys travel time between food patches. Taken together, our results reveal a neuronal substrate that integrates nutritional state and patch density information to control a specific foraging decision, and therefore provide an important step towards a mechanistic explanation of the cognitive computations that resolve complex cost-benefit trade-offs.