A complex assessment of animal behavior in a learning paradigm

L.Yu. Rizhova¹, O.G. Semionova² and V.G. Shaliapina¹

Laboratory of Neuroendocrinology, Pavlov Institute of Physiology of Russian Academy of Sciences, sankt-Petersburg, Russia.

Evaluation of the effects of any treatment on animal learning is a popular pursuit in modern neuroscience. Since numerous cognitive and non-cognitive functions affect performance in a learning task, it is always dif.cult to determine what process was changed by the treatment, and improved or impaired learning. We share the view that measuring multiple behavioral parameters in a single testing paradigm can often be a valid way to answer this question. The test should be designed so that the effects of confounding factors such as lateral bias, anxiety, positive or negative reinforcement, and high or low test animal motor activity are minimized. Accordingly, we adapted the T-maze test for the assessment of learning in rats. To avoid handling of animals in the maze, we constructed doors at both ends of the maze. A dark transport cage was attached to each door. The animal was placed in the stem and given five minutes of adaptation. Pellets of cheese were placed at both ends and the doors were closed. During this time the following behavioral parameters were registered: latency (a period before the rat began to enter an alley), rearing, locomotion, grooming, eating, defecation. Animals were then taught to go to the alley opposite the one chosen at the first trial. This time the originally chosen alley did not contain any cheese, and its exit was always closed. The food was placed near the opened exit of the other alley, and in the attached transport cage. Each animal had to go through this procedure several times, until it made four correct turns in succession. The number of trials necessary to achieve this was taken as a measure of learning. Some animals did not exhibit interest in food during the learning procedure. Escape from the maze served as their reinforcement for learning. We segregated eating/not eating animals from the experimental and control groups and compared their learning abilities. The same procedure could be followed for the animals with short or long latency, high and low locomotion activity, strong or weak lateral bias, etc. Thus we reveal the processes affected by the treatment, and which by the way influenced learning. This method be also used for integral evaluation of animal anxiety.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.

© 2005 Noldus Information Technology bv