Behavioral Strategies and Neural Mechanisms of Dynamic Foraging

Abstract

In dynamic environments, humans and animals constantly use reward and error signals to guide action selection. As diverse strategies may be used in the task, it is unclear how neural processing of reward, value and action information might change across different behavioral states. This thesis develops techniques for behavioral and neural dissection of cortical contributions to dynamic foraging behavior. By a combination of state-space behavioral modeling, large-scale widefield and single-cell imaging technologies, together with encoding models of neural responses, we showed and quantified the rich behavioral repertoire of rodents, state-dependent cortical processing of trial outcome, and diverse, distributed encoding of value and action switching information at the single-cell level.

Analysis of rodent behavior revealed mixtures of behavioral modes during dynamic foraging, characterized by different switch offsets, sharpness and exploration rates. We developed a new computational approach, block Hidden Markov Model, to characterize and identify these discrete states of behavior in blocks of trials. These states can be accurately decoded as sub-regimes of model-free or inference-based behavior. Widefield imaging and unsupervised analysis of the cortical activity during the behavior revealed distinct cortical activation modes corresponding to the frontal, motor, visual and retrosplenial regions that have different dynamic representations of rewards and errors. Dissecting single-neuron responses in these candidate regions with three-photon imaging across cortical depths revealed specialized processing of reward-related variables. We found widespread representation of outcome, value and switching in all four regions, with an enriched representation of outcome in the retrosplenial cortex (RSC), and of action values in the anterior cingulate cortex (ACC). Using block Hidden Markov Model, we found outcome representation was enhanced in high-efficiency behavioral states but only weakly represented at the low-efficiency states. Optogenetic perturbations of outcome information in the frontal and RSC neural clusters decreased the frequency of these high-efficiency states, demonstrating their causal role in the expression of efficient switching behavior.

Together, these computational methods and experiments provided new tools for quantification of dynamic foraging behavior, and revealed specialized and state-dependent distribution of outcome, value and switch representations in key cortical regions. These insights lead to important hypotheses about cortical-subcortical interactions during reward-guided behavior.