The behavioral microstructure of a memory-guided food-caching behavior and its relationship to hippocampal replay

Project Summary The hippocampus is a critical site for rapid memory formation and retrieval, with extensively documented functions representing spatial and navigational variables, yet less is known of the means by which it guides behavior. Bursts of hippocampal activity identified in many species as sharp-wave ripple (SWR) events are promising neural mechanisms to link hippocampal activity and behavioral function, since during SWRs hippocampus can represent positions the animal previously experienced, or will traverse in the near future. This proposal considers the food-caching behaviors of chickadees, who use a hippocampal-dependent, one- shot memory to guide cache retrieval. This behavior is exceptionally promising to link hippocampal activity to individual experiences, which has been a barrier to determining the nature of representations during SWR and their function in previous studies. These studies will develop a comprehensive account of how chickadee behavior is affected by cache site memory, determine whether SWRs are related to memory-specific components of caching behavior, and determine whether SWRs represent individual cache sites to assist later retrieval. I hypothesize that memory-guided behavior is distinguishable from random foraging. I will use video tracking to measure the behavioral microstructure of chickadee caching behavior, including postural and gaze time series, and develop quantitative models of behavioral motifs and their temporal structure. Using these tools to compare memory-guided and random foraging behaviors, I will develop decoders that indicate precisely how and when behavior appears memory-guided. I hypothesize that SWRs occur more often, and are more predictive of behavior, when behavior is memory- guided. I will use electrophysiology in freely moving chickadees during caching to relate SWRs to the microstructural analysis of behavior and decoding of behavioral strategy. I hypothesize that SWRs reactivate representations of previous, individual caching events, and that this reactivation improves subsequent retrieval for that individual site. Using electrophysiological recordings in caching birds, I will relate reactivations to the accuracy of retrieval behavior. Using online SWR detection and suppression, I will further determine if SWRs causally affect later retrieval behavior.