ePoster

Environmental complexity modulates the arbitration between deliberative and habitual decision-making

Ugurcan Mugan,Samantha Hoffman,Paul J. Cunningham,Paul S. Regier,Seiichiro Amemiya,A. David Redish
COSYNE 2022(2022)
Lisbon, Portugal
Presented: Mar 18, 2022

Conference

COSYNE 2022

Lisbon, Portugal

Resources

Authors & Affiliations

Ugurcan Mugan,Samantha Hoffman,Paul J. Cunningham,Paul S. Regier,Seiichiro Amemiya,A. David Redish

Abstract

Research into decision-making suggest that the brain contains multiple systems for generating adaptive strategies. The usual explanation is that each strategy is optimized for different environmental conditions. Neurophysiologically, current theories suggest that habitual control requires dorsolateral striatal (DLS) networks, while deliberative control depends on hippocampus (HC) and medial prefrontal cortex (mPFC). Within mPFC, manipulation studies and prior research suggest that prelimbic cortex (PL) contains strategy representations making it especially important for this interaction. To examine the role of environmental structure on the arbitration between decision-making systems, we first re-analyzed behavior and recordings from HC and DLS during a contingency-switching task in the light of a novel graph-theoretic measure of maze complexity. Second, we chemogenetically disrupted PL while explicitly controlling for maze complexity. In HC, both the duration of the ascending phase of the theta cycles---hypothesized to contain more nonlocal representations---and overall rate and duration of sharp wave ripple events increased with environmental complexity. In DLS, the development of the preferential firing at the start and end of a lap---task-bracketing---increased with environmental simplicity. PL inactivation reduced deliberation and led to deficits in performance, particularly in complex environments. Furthermore, it resulted in an increase in behavioral stereotypy and a decrease in spatial exploration, but again, only in complex environments. Conversely, in simple environments, we found no significant difference in the durations of the decision-making mode or overall exploration. This data suggests that mPFC engagement---hypothesized to be important for both strategy-setting and the initiation of deliberative sequences---may be particularly crucial in complex environments. Consistent with current theories, more complex environments likely require a higher cognitive memory load resulting in slower automation. PFC disruption may cause working memory disruptions that were consequential for performance in complex environments in which deliberation is necessary.

Unique ID: cosyne-22/environmental-complexity-modulates-e105c5a8