Prior actions influence cost-benefit-related decision-making during mouse foraging behaviours

Food foraging is essential for the fitness of animals. Previous studies have suggested that optimal foraging strategies involve a cost-benefit analysis comparing reward versus effort to guide action choices. Little is known how prior experience with different actions to obtain rewards may affect subsequent foraging choices. Here, we report a sunflower seed foraging test to investigate how effort and prior actions influence decision-making in laboratory mice. Sunflower seeds are a natural food favourite for mice, and mice spend effort to peel the hard shells to obtain the seeds. In our test, peeled and unpeeled sunflower seeds were placed at different ends of a Y-maze. Mice were free to explore the maze and make foraging decisions. Naive mice were more likely to choose peeled seeds requiring low effort versus unpeeled seeds requiring high effort. Furthermore, mice with prior seed peeling experience significantly reduced preference for peeled seeds during the subsequent Y-maze foraging test, compared with mice pre-exposed to peeled seeds only. This experience-dependent shift in foraging choice was associated with reduced seed peeling time and improved motor skills with practice, and predictable on a trial-by-trial basis by a probabilistic decision-making model with the amount of peeled and unpeeled seeds consumed as inputs. Together, these results suggest that laboratory mice make rational foraging choices based on effort estimation and moreover, prior actions to obtain reward alter effort estimation and decision-making through motor skill learning. This naturalist behavioural task may be applied to dissect neural mechanisms in adaptive decision-making during foraging.

Automated summary

Prior actions can affect foraging choices in animals. Laboratory mice make rational foraging choices based on effort estimation, and prior actions to obtain rewards alter effort estimation and decision-making through motor skill learning.