Learning Strategies and Corresponding Brain Region in Navigation Tasks
When animals migrate to a new environment, they need to learn how to find the food to survive. It is very similar that human need to learn the driving route when they travel and thus be able to navigate to destinations. The navigation strategies for positive reward in an unfamiliar environment have been studied throughout history. Early pioneer studies by Pavlov (1927) suggested that there was a stimulus-response (S-R) habit formation involved in animal learning processes. His famous classical conditioning studies showed that the responses could automatically initiate without stimulus after overstimulation. In order to better understand this behavior pattern, the transition from the simple motor movement to the S-R habit thus has been studied in animal models (Galanter & Shaw, 1954; Tolman, Ritchie, & Kalish, 1946). Two types of learning have been identified in maze-training of rats. The first strategy is referred as “place strategy”, by which rats select their paths to the rewards based on their own physical locations. The second strategy, based on S-R habit model, is called “response learning”, by which rats tend to follow their acquired motor sequence regardless of the locations of the rewards. These experiments (Galanter & Shaw, 1954; Tolman, Ritchie, & Kalish, 1946) suggest that the place strategy is preferred for significant portion of subjects, learned faster than response strategy, and has less error during the early stage of navigation training in maze; whereas the response strategy dominates the later stage.
Based on pilot studies that theorized the learning strategies in navigation, recent scientists were more focused on the neural basis of place and response learning. According to the research of food searching tasks for animals (Compton, 2004; Lee, & Kim, 2010; Robinson, Sotak, During, & Palmiter, 2006) and virtual navigation tasks for humans (Schmitzer-Torbert, 2007; Bohbot, Gupta, Banner, & Dahmani, 2011), two brain areas have been found to be responsible for place and response learning. On the one hand, the hippocampus has been found to be the critical region to facilitate place strategies. Findings confirm that the neural activities in hippocampus are more active during the early stage of training. On the other hand, response learning requires the involvement of the basal ganglia.
However, not all the studies of the neural circuits are considered to be convincing. Among these findings, the rats-based experiments are more reliable and practical because human models are limited in many ways. For example, the response strategy acquisition can be facilitated by the inhibition of hippocampus in rats (Packard & McGaugh, 1996), but such a procedure is not practical for human subjects. Brain lesion and other cerebral manipulations are rarely found in human-based research compared with animal models. Some human-based studies, however, have recruited fMRI recordings for the compensation of inability to make direct cerebral manipulation (de Wit, Aitken, Dickinson, & Fletcher, 2009; Sadeh, Shohamy, Levy, Reggev, & Maril, 2011; Tricomi, Balleine, & O'Doherty, 2009). Thus the shortages of human-based research are that one cannot manipulate to neurological structures (e.g. hippocampus, basal ganglia) to allow strong causal inferences between brain activities and learning strategies. Fortunately, the findings from rat-models are promising and maintain high internal validity and often provide general guidance for designing human-based experiments.
Based on the learning strategy theories and their correspondent brain regions, this literature review will cover recent findings in both animal-based and human-based research about learning strategies involved in navigation tasks. Then hopefully this article can show the connections between these two domains