Sensorimotor dynamics in the superior colliculus of the echolocating bat

Abstract

To successfully execute natural tasks, animals must continuously monitor and integrate dynamic sensory information to guide behavior. While this process - known as active sensing - is fundamental to goal-directed behavior, the mechanisms by which sensory information supports motor planning, particularly across different behavioral contexts, remains poorly understood. We investigated sensory and motor activity in the superior colliculus (SC) of echolocating bats, a powerful model system to study active sensing behaviors due to their reliance on self-generated sonar signals. We recorded multichannel SC activity while bats performed spatial navigation and target tracking behaviors. We hypothesized that changes in sensory signals (i.e., returning echoes) drive quantifiable adjustments in SC activity related to sonar call production, and that adaptive changes in SC sensorimotor processing are preserved across behavioral contexts. We examined pooled SC activity using a dimensionality reduction technique to isolate the largest changes in firing rates across multichannel SC recordings. We then examined the variation in SC activity from the time of echo arrival (sensory input) to call production (motor output). Our data show that increases in call rate are associated with shorter trajectory lengths in neural state space, reflecting reduced variability and increased efficiency in population activity. Strikingly, when bats cluster calls into sonar sound groups (SSGs, which are rapid bursts of calls to increase sensory sampling), SC activity is distinct from times when bats produce isolated calls. Notably, we find that successive SSGs lead to progressively shorter trajectory lengths, suggesting a refinement of sensorimotor processing across successive SSGs. Together, our findings demonstrate that sensorimotor population dynamics in the SC follow common principles across distinct behavioral tasks. These results support a fundamental role for the SC in adaptive sensing and highlight its contribution to flexible sensorimotor tranformations in natural behaviors.