Transition Dynamics in Spatial Choice

Abstract

Previous research has provided abundant evidence that negative as well as positive reinforcers can elicit operant responding in rats. Less is known, however, about the relative impact of positive versus negative reinforcers on the dynamics of spatial choice in a complex setting, including old and new stimuli. To study such dynamics, we devised a behavioral paradigm consisting of an initial learning phase, followed by a “transition test,” which included a novel stimulus. In the initial learning phase, rats acquired an operant nose-poke response to a visual (LED) onset at the central hole in the front wall of a Skinner box. We trained two groups of rats, with different types of reinforcement; Group 1(appetitive learning) obtained a food pellet as reinforcement for correct responding, whereas Group 2 (appetitive and avoidance learning) obtained a food pellet and avoided a foot shock as reinforcement for correct responding. Once both groups had established similar, near-perfect levels of performance, we introduced a first transition test, in which a second visual (LED) onset was presented simultaneously with the visual onset at the center. The rats were free to continue responding to the central hole (with the same reinforcement schedule for central responding as during the initial learning). However, the rats obtained three instead of only one food pellet for responding to the peripheral hole (Group 1 and Group 2) as well as avoiding a foot shock (Group 2). Overall, as compared to rats in Group 1, the rats that underwent both appetitive and avoidance learning showed more efficient transition, switching to peripheral responding. These data cannot easily be accommodated by operant conditioning models, according to which the rats in Group 2 should have been particularly motivated to continue responding at the central hole. Instead, we propose that their spatial choice was modified by a complex interaction of classic and operant learning.