Extrinsic inputs exert generalized actions on the feeding CPG of Aplysia

All animals, from simple invertebrates to humans, live in complex environments in which they must respond differentially to different stimuli. Responses to different stimuli often require the efficient execution of complex motor behaviors. Complex motor behaviors can share the same core movements while generating radically different behaviors. Often times, complex motor behaviors are controlled by a single neural network that generates the shared and unshared movements between different behaviors. The broad aim of my research was to understand the neural mechanisms that behavior-generating networks exploit in order to efficiently respond to stimuli that are behaviorally related.;An important aspect of generating appropriate and proficient behaviors is to respond preferentially to persistent stimuli and suppress responses to transient stimuli. In this end, behavior generating networks must "remember" previous experiences. This means that an animal must respond differently in response to the same stimulus depending on its recent experience. Thus, a simple stimulus-response relationship does not occur within neural networks. Instead, persistent stimuli establish a network "state" that is based on the previous activity of the network and can influence the response to subsequent stimuli.;I wanted to determine how network states and the mechanisms that establish them influence the shared movements underlying related but distinct behaviors. As these behaviors are related, they in turn share some of the same movements. Thus, persistent activity of one behavior could beneficially influence the other behavior. Here, I used a simple invertebrate behavior-generating network as it is difficult to determine the neuronal mechanisms controlling network states in higher-order systems. I studied the properties of the neuronal network that mediates feeding responses in the sea slug Aplysia californica. A major conclusion of this dissertation is that the state of the feeding neural network depends on its previous history of activity. With respect to shared movements, I found that activation of different feeding inputs established a convergent network state that allowed the network to respond similarly to different stimulus presentations. Additionally, the mechanisms underlying the establishment of this network state were similar between the related but distinct behaviors.