Abstract
Neural circuits possess a dynamic flexibility to adapt to changing inputs and integrate these experiences into their functional framework to allow an organism to learn, adapt, and survive. This encoding is tightly regulated by the interplay of different mechanisms that enable plasticity, while maintaining the overall stability of the system. Hebbian plasticity mechanisms drive changes in synaptic strengths and firing rates, but if left unchecked, can result in runaway activity and lead to aberrant circuit function. Therefore, homeostatic plasticity mechanisms serve to constrain these changes and provide a regulatory force on neural activity. The firing rates of cortical neurons are thought to be subject to these mechanisms: individual firing rates are stable across their distribution position, time, light-dark transitions, and even in the face of dramatic sensory perturbations. However, how firing rates are affected by more salient drivers of activity, such as learning, had not be observed in vivo. In this thesis, I challenge this idea by asking how naturalistic, visually guided, prey capture learning changes V1 activity. I employed a combination of behavioral analysis, chemogenetics, chronic in vivo electrophysiology, slice electrophysiology, and pharmacology to measure the effects of learning on primary visual cortex (V1) and probe the role of several homeostatic plasticity mechanisms in stabilizing those changes. I show that juvenile rats can rapidly learn how to hunt, capture, and consume crickets in a V1-dependent manner, which drives the recruitment of pursuit-tuned neurons across learning. I also present evidence that prey capture, an ethological learning paradigm, can drive persistent changes in their firing rates and dendritic spine number, suggesting learning resets the setpoints of primary binocular visual cortex neurons (V1b). I also argue that these changes are gated by homeostatic mechanisms, due to their slow timescale, gating by behavioral state, and dependence on TNFα signaling, and are important for future behavioral performance. In summary, this thesis explores how V1 activity setpoints are malleable when driven by ethological means and begins to explore how homeostatic mechanisms gate this reconfiguration of V1 circuitry.