Sequential behavior in the rat: a new model using food-reinforced instrumental behavior.

Sequential behavior, probably reflecting procedural learning, has intensively been investigated in humans. This work has mainly been done using so-called serial reaction time tasks. In such tasks, subjects have to respond rapidly to simple visual stimuli appearing at one of four locations by pressing a corresponding response key. Unknown to the subjects, these stimuli can follow a specific repeating sequence. Learning of such a sequence is typically inferred from faster reaction times to sequence as compared to random blocks of stimuli. In contrast to human subjects, the analysis of sequential behavior has received considerably less attention in rodents, possibly due to the lack of analogous animal models there. In order to establish such a model, a method was developed in rats to investigate serial reactions under conditions of random or sequential stimulus presentation. Operant testing chambers were used which consisted of four nose-poke holes with cue lights. These holes were arranged in a square fashion with a pellet receptacle in the center. The task of the rat was to rapidly respond to an illuminated hole by poking into it in order to obtain food. The stimulus locations varied permanently, and these changes pursued either a random or serial order. In three experiments with differing methodological details, responding under such conditions was analyzed with sequences consisting of 6, 12 or 13 positions. Evidence was obtained that rats can improve their performance under sequence as compared to random conditions, for example, with respect to the percentage of reinforcements obtained, or with respect to reaction times. Furthermore, methodological factors, like response requirements, were addressed which may critically affect experimental outcome. Together, this new kind of instrumental task might be useful to analyze sequential performance in the rat, and the brain mechanisms by which it is mediated.